Oracle Communications

Diameter Signaling Router

Cloud Benchmarking Guide
Release 8.4.x
February 2020
Table of Contents
Introduction .................................................................................................................................6
References .................................................................................................................................6
Acronyms ....................................................................................................................................6
Terminology ................................................................................................................................8
About Cloud Deployable DSR .....................................................................................................9
What is a Cloud Deployable DSR? ........................................................................................9
Infrastructure Matters ..........................................................................................................10
Flexibility .............................................................................................................................10
Methodology .......................................................................................................................10
Benchmarking Cloud Deployable DSR ......................................................................................11
What to do with all this data?...............................................................................................11
Infrastructure Environment ........................................................................................................12
General Rules for All Infrastructures....................................................................................13
Hyper-Threading and CPU Over-Subscription ...............................................................13
VM Packing Rate ...........................................................................................................14
Infrastructure Tuning .....................................................................................................14
KVM (QEMU)/Oracle X5-2 – Infrastructure Environment .....................................................14
Device Drivers ...............................................................................................................15
BIOS Power Settings .....................................................................................................15
Guest Caching Modes ...................................................................................................15
Memory Tuning Parameters ..........................................................................................16
VMware (ESXi) – Infrastructure Environment ......................................................................17
Virtual Sockets vs. Cores per Virtual Socket .................................................................18
Network Settings ...........................................................................................................18
Power Settings ..............................................................................................................19
Hardware Assisted Virtualization ...................................................................................19
Virtual Machine Storage Configuration ..........................................................................19
Disk Provisioning ...........................................................................................................20
Benchmark Testing ...................................................................................................................20
DA MP Relay Benchmark .........................................................................................................20
Overview .............................................................................................................................20
Topology .............................................................................................................................21
Message Flow .....................................................................................................................21
Indicative Alarms/Events .....................................................................................................21
DA-MP Indicators for Capacity Expansion ...........................................................................22
DA-MP Utilization ..........................................................................................................22
Measuring DA-MP Connection Utilization ......................................................................23
Suggested Resolution .........................................................................................................23
Full Address Based Resolution (FABR) Capacity ......................................................................23
Overview .............................................................................................................................23
Topology .............................................................................................................................24
SDS DB Details .............................................................................................................24
Message Flow .....................................................................................................................24

DSR 8.4 Cloud Benchmarking Guide 2

 DP Indicators for Capacity Expansion .................................................................................25
Measuring DP Utilization ...............................................................................................25
Suggested Resolution .........................................................................................................26
SS7 MP.....................................................................................................................................26
Overview .............................................................................................................................26
Topology .............................................................................................................................27
Message Flow .....................................................................................................................27
Indicative Alarms/Events .....................................................................................................27
SS7 MP Indicators for Capacity Expansion .........................................................................28
Measuring SS7 MP Utilization .......................................................................................28
Suggested Resolution .........................................................................................................28
vSTP MP...................................................................................................................................28
Overview .............................................................................................................................28
vSTP MP Indicators for Capacity Expansion .......................................................................29
Utilization .......................................................................................................................29
Suggested Resolution .........................................................................................................29
Policy DRA (PDRA) Benchmarking ...........................................................................................29
Overview .............................................................................................................................29
Topology .............................................................................................................................30
Message Flow .....................................................................................................................31
Indicative Alarms/Events .....................................................................................................32
Measurements ....................................................................................................................33
Suggested Resolution .........................................................................................................33
Diameter Security Application (DSA) Benchmarking .................................................................34
Topology .............................................................................................................................34
Service Capability Exposure Function (SCEF) Benchmarking...................................................35
Topology .............................................................................................................................35
Suggested Resolution .........................................................................................................36
Equipment Identity Register (EIR) .............................................................................................36
NOAM .......................................................................................................................................37
Overview .............................................................................................................................37
Indicative Alarms/Events .....................................................................................................37
Measurements ....................................................................................................................37
Measuring Network OAM Utilization ..............................................................................37
Suggested Resolution .........................................................................................................37
SOAM .......................................................................................................................................38
Overview .............................................................................................................................38
Indicative Alarms/Events .....................................................................................................38
Suggested Resolution .........................................................................................................38
IPFE..........................................................................................................................................38
Overview .............................................................................................................................38
Topology .............................................................................................................................39
Indicative Alarms/Events .....................................................................................................39
Suggested Resolution .........................................................................................................39

DSR 8.4 Cloud Benchmarking Guide 3

IDIH ..........................................................................................................................................40
Overview .............................................................................................................................40
Suggested Resolution .........................................................................................................40
Appendix A. DSR VM Configurations ........................................................................................40
Appendix B. DSR VM Disk Requirements .................................................................................43
Appendix C. VM Networking Requirements ..............................................................................44
Appendix D. Summary of Benchmark Data Recommendations.................................................46
Appendix E. Detailed Infrastructure Settings .............................................................................49
Appendix F. Small DSR VM Configuration ................................................................................50
Appendix G. DSR DA-MP Profiles.............................................................................................51
Infrastructure Performance Differences ...............................................................................51
Traffic Mix ...........................................................................................................................51
VM Configuration ................................................................................................................52
Selecting a DA-MP Traffic Profile ........................................................................................52

List of Tables
Table 1: Acronyms ....................................................................................................................................... 6
Table 2: Terminology ................................................................................................................................... 8
Table 3: Virtual Network Interrupt Coalescing and SplitRX Mode ............................................................. 18
Table 4: Power Management Profiles ........................................................................................................ 19
Table 5: Virtualization Performance by Processor ..................................................................................... 19
Table 6: DA-MP Test Set Up ..................................................................................................................... 21
Table 7: DA-MP Alarms/Events ................................................................................................................. 21
Table 8: DA-MP Utilization Metrics ............................................................................................................ 22
Table 9: DA-MP Connection Metrics .......................................................................................................... 23
Table 10: SDS DP Message Details .......................................................................................................... 24
Table 11: SDS DP Alarms/Events ............................................................................................................. 25
Table 12: SDS DP Utilization Metrics ........................................................................................................ 25
Table 13: DP SOAM Metrics ...................................................................................................................... 26
Table 14: SS7 MP Message Detail ............................................................................................................ 27
Table 15: SS7 MP Alarms/Events .............................................................................................................. 27
Table 16: SS7 MP Metrics ......................................................................................................................... 28
Table 17: VSTP-MP Utilization Metrics ...................................................................................................... 29
Table 18: PDRA Test Call Model ............................................................................................................... 32
Table 19: SBR (b) Alarms/Events .............................................................................................................. 32
Table 20: Session SBR (b) VMs Metrics .................................................................................................... 33
Table 21: Binding SBR (b) Server Metrics ................................................................................................. 33
Table 22: Stateful and Statelss Counter Measures ................................................................................... 34
Table 23: 10K MPS DA-MP VM Profile ...................................................................................................... 35

DSR 8.4 Cloud Benchmarking Guide 4

Table 24: SCEF DA-MP Server Metrics ..................................................................................................... 36
Table 25: Network OAM Metrics ................................................................................................................ 37
Table 26: System OAM Metrics ................................................................................................................. 38
Table 27: IPFE Throughput ........................................................................................................................ 38
Table 28: IPFE Metrics ............................................................................................................................... 39
Table 29: VM Configurations and Affinity Rules ........................................................................................ 40
Table 30: VM Disk Utilization Characteristics ............................................................................................ 43
Table 31: VM Networking Utilization Characteristics ................................................................................. 44
Table 32: Typical OCDSR Network to Device Assignments ...................................................................... 45
Table 33: Benchmark Data Summary ........................................................................................................ 46
Table 34: Recommended Maximum Engineering Targets ........................................................................ 46
Table 35: Detailed Infrastructure Settings .................................................................................................. 49
Table 36: Minimal System VM Configurations and Affinity Rules .............................................................. 50

List of Figures
Figure 1: DA-MP Testing Topology ............................................................................................................ 21
Figure 2: DA-MP Message Sequence ....................................................................................................... 21
Figure 3: SDS DP Testing Topology .......................................................................................................... 24
Figure 4: SDS DP Message Sequence ...................................................................................................... 24
Figure 5: SS7 MP Testing Topology .......................................................................................................... 27
Figure 6: SS7 MP Message Flow .............................................................................................................. 27
Figure 7: SBR Testing Topology ................................................................................................................ 30
Figure 8: PDRA Message Sequence ......................................................................................................... 31
Figure 9: DSA Testing Topology ................................................................................................................ 34
Figure 10: SCEF Testing Topology ............................................................................................................ 36
Figure 11: EIR Testing Topology ............................................................................................................... 37
Figure 12. IPFE on Ingress Side Only ....................................................................................................... 39
Figure 13: IPFE on both Ingress and Egress Sides ................................................................................... 39

DSR 8.4 Cloud Benchmarking Guide 5

Introduction
The Oracle Communications Diameter Signaling Router (OCDSR or DSR) is deployable in the cloud as a
Virtual Network Function (VNF). With DSR’s added flexibility of being cloud deployable, operators must
be able to manage the capacity and performance of the DSR in the cloud.
This document focuses on:
• How to benchmark DSR performance and capacity in a cloud deployed DSR
• Provides recommendations on performance tuning the DSR
• Provides benchmark data from our labs
• Provides information on the key metrics used to manage DSR performance and capacity
• Provides recommendations on how to use the data obtained from the metrics

References
[1] Performance Best Practices for VMware vSphere® 6.0 at
http://www.vmware.com/content/dam/digitalmarketing/vmware/en/pdf/techpaper/vmware-perfbest-
practices-vsphere6-0-white-paper.pdf
The following are available at Oracle.com on the Oracle Help Center (OHC):
[2] DSR Alarms, KPIs, and Measurements
[3] DSR Cloud Deployable Installation Guide
[4] Policy and Charging Application User’s Guide

Acronyms
Table 1: Acronyms
Acronym Description
API Application Programming Interface
ARR Application Route Rule
ART Application Route Table
CM Counter Measure
COTS Commercial Off The Shelf
CPU Central Processing Unit
DA-MP Diameter Agent Message Processor
DB Database
DP Database Processor
DSA Diameter Security Application
DSR Diameter Signaling Router
EIR Equipment Identity Register
ETG Egress Throttle Group
FABR Full Address Based Resolution
FQDN Fully Qualified Domain Name
GB Gigabyte
HDD Hard Disk Drive

DSR 8.4 Cloud Benchmarking Guide 6

 Acronym Description
HP Hewlett Packard
HSS Home Subscriber Server
HTTP Hypertext Transfer Protocol
ID Identification
IDIH Integrated Diameter Intelligence Hub
IMI Internal Message Interface
IMSI International Mobile Subscriber Identity
I/O Input/Output
IOP Interoperability
IoT Internet of Things
IP Internet Protocol
IPCAN Internet Protocol Connectivity Access Network
IPFE IP Front End
KPI Key Performance Indicator
KSM Kernel Same-page Merging
KVM Kernel-based virtual machine
LSI Large Scale Integration
LTE Long Term Evolution
MD-IWF MAP-Diameter Interworking Function
MME Mobility Management Entity
MNO Mobile Network Operator
MP Message Processor
MPS Messages Per Second
MSISDN Mobile Station International Subscriber Directory Number
MTC Machine Type Communication
NIC Network Interface Card
NOAM Network Operations, Alarms, Measurements
NE Network Element
OAM Operations, Administration, and Maintenance
OCDSR Oracle Communications Diameter Signaling Router
OCSG Oracle Communications Services Gatekeeper
OHC Oracle Help Center
PCRF Policy and Charging Rules Function
PDRA Policy Diameter Routing Agent
PRR Peer Route Rule
PVSCSI Paravirtual SCSI
RAM Random Access Memory
RBAR Range Based Address Resolution
SAS Serial Attached SCSC

DSR 8.4 Cloud Benchmarking Guide 7

 Acronym Description
SCEF Service Capabilities Exposure Function
SBR Session Binding Repository
SBR(b) SBR – subscriber binding database
SBR-g
SBR(s) SBR – session database
SBR(u) SBR – universal database
SCS/AS Service Centralization and Continuity Application Server
SCSI Small Computer System Interface
SDS Subscriber Database Server
SFF Small Form Factor
SGSN Serving GPRS Support Node
SMS Short Message Service
SOAM System (nodal) Operations, Alarms, Measurements
SS7 Signaling System #7
SSD Solid State Drive
THP Transparent Huge Pages
TSA Target Set Address
TTP Troubleshooting Trigger Point
vSTP virtual Signaling Transfer Point
UDR-NO User Data Repository Network OAM & Provisioning
U-SBR Universal SBR
VM Virtual Machine
VNF Virtual Network Function
VoLTE Voice over LTE
WAN Wide Area Network
XMI External Management Interface
XSI External Signaling Interface

Terminology
Table 2: Terminology
Term Description
1+1 For every 1, an additional 1 is needed to support redundant capacity. The specific
Redundancy redundancy scheme is not inferred (for example, active-active, and active-standby).
Geo-Diverse Refers to DSR equipment located at geographically separated sites/datacenters
Geo- A node at a geo-diverse location which can assume the processing load for another
Redundant DSR signaling node(s)
Ingress A measure of the total Diameter messages per second ingressing the DSR. For this
Message measure, a message is defined as any Diameter message that DSR reads from a
Rate Diameter peer connection independent of how the message is processed by the DSR.

DSR 8.4 Cloud Benchmarking Guide 8

 Term Description
Messages A measure of the DSR Diameter message processing volume in messages per second.
Per Second For this measure, a message is defined as:
• DSR processing of an ingress Diameter message and either transmitting a single
outgoing Diameter message or discarding the ingress message. The outgoing
message may be a variant of, or a response to, the ingress message.
• DSR transmission of any Diameter message, as required by DSR configuration,
that is associated with incremental actions/events associated with #1 above. For
example, the re-routing of a Request upon connection failure or the copying of a
Request.
Messages excluded from this measure are:
• Diameter peer-to-peer messages: CER/CEA, DWR/DWA, and DPR/DPA
• Ingress Diameter messages discarded by the DSR due to Overload controls
• Answers received in response to Message Copy
For the vSTP MP the MPS excludes the equivalent SSNM status management
messages.
N+K For every N, an additional K is needed to support redundant capacity. The specific
Redundancy redundancy scheme is not inferred (for example, active-active, active-standby).
Node A DSR node is a DSR signaling node (SOAM and subtending topology), an NOAM
node or an SDS node. A node is synonymous with the network element (NE).
Site A specific geographic location or datacenter where DSR application is installed.

About Cloud Deployable DSR

DSR is deployed on a number of platforms. The DSR has a multiple deployment scenarios:
• Bare-metal and hybrid (mixture of bare metal and virtual machines) — is the original deployment
configuration of the DSR. It scales to very high performance and is widely deployed.
• Fully virtualized — was introduced shortly after bare-metal. It provides virtualization of the DSR, but
does not use a cloud manager, and does not co-reside with other applications. Provides a compact,
cost-effective footprint and is widely deployed.
• Cloud deployable – It provides full virtualization, assumes the DSR resources are managed by a
COTS cloud manager, and the DSR can be one of many applications in the cloud. Cloud deployable
DSR is the focus of this document.
• Mix and match – DSR is a network of DSR signaling sites. The deployment infrastructure at each site
can vary, for example, bare-metal at one site, and then cloud deployed at another location.

What is a Cloud Deployable DSR?

A DSR that is ready and able to be deployed into a number of different cloud environments, including but
not limited to:
• A customer provided cloud infrastructure. The DSR is simply one of many applications.
• A dedicated private cloud. The DSR may be the only application, or one of a small set of
applications. Services and infrastructure may also be provided by Oracle and deployed at customer’s
sites. Often (but not necessarily) this is a deployment tuned specifically for the DSR.
• A hosted cloud. The DSR is deployed in an Oracle or operator hosting cloud, and end-customers
rent or lease the DSR application from the hosting provider.

DSR 8.4 Cloud Benchmarking Guide 9

Infrastructure Matters
The DSR is capable of running on a huge variety of infrastructures, but not all infrastructures are the
same. The performance, capacity, and latency of the DSR can vary dramatically based on the chosen
infrastructure and how it is deployed. In general, the DSR works best in a high bandwidth, low-latency,
high processing power environment (carrier grade cloud). Some considerations that impact DSR
performance, capacity, latency:
• Hardware – the CPUs and NICs (network interface cards)
• Hypervisor settings/configuration
• Uplinks, switches, WAN latency
• Storage configuration (local, networked)
DSR has excellent high availability and geo-diversity resiliency mechanisms that work in concert with
cloud manager capabilities. Obviously the needed scale, availability, and resiliency of the deployment
also impact the resource and infrastructure requirements.

Flexibility
DSR is flexibly deployed into many different clouds. It is unlikely that any two clouds are exactly the
same and operators need to optimize for different reasons (for example, power consumption may be
critical for one operator, and WAN latency at another), varying sets of applications, and differing
operational requirements. The performance and capacity of the DSR varies in each cloud, and the DSR
application can no longer provide a guaranteed level of performance and capacity. However, the
operator still needs to:
• Plan their networks – DSRs use resources, what impact DSR has on their datacenters?
• Deploy DSR with predictable (if not exact) performance and capacity.
• Manage the capacity and performance of the DSR in their datacenters.

Methodology
There is a set of DSR specific tools, methods and documentation to assist in planning, deploying, and
managing the capacity and performance of a cloud deployable DSR. This toolset is expected to be used
in conjunction with information and tools provided by the infrastructure (hardware, cloud manager,
hypervisor) vendors.
• Planning for cloud deployable DSR
• Estimating required resources for a given DSR cloud deployment
• Please contact your Oracle Sales Consultant. They have access to the DSR Cloud
Dimensioning tool which estimates DSR cloud resources. This tool takes into account many
factors not covered in this benchmarking guide, such as the overhead for optional DSR
features not covered in the benchmarking guide, and recommended margins for redundancy.
• DSR Cloud Customer Documentation
• Can be found with the DSR customer documentation at Oracle.com on the Oracle Help
Center (OHC).
• Look under the topic: “Cloud Planning, Installation, Upgrade, and Disaster Recovery.”
• Deploy DSR with predictable performance and capacity
• It is recommended that the DSR is run through a benchmark on the target cloud infrastructure to
determine the likely capacity and performance in the target infrastructure. This information can

DSR 8.4 Cloud Benchmarking Guide 10

 then be used to adjust the initial deployment resources (if needed), and to help predict future
resource requirements if and when the DSR grows.
• This document provides information on how to benchmark DSR performance and capacity. It
also provides comprehensive benchmark results for a few select infrastructures. More
benchmark data will be added to the document as it becomes available.
• This document also provides performance recommendations and observed differences for
performance tuning decisions.
• Manage the capacity and performance of the DSR
• The customer network is always changing- traffic patterns change, new applications are
introduced. The infrastructure changes – new hardware, software/firmware updates. The
operator needs to monitor and adjust the DSR resources for the changing conditions of the
network and infrastructure.
• This document provides the key metrics and recommendations for monitoring the capacity and
performance of a cloud deployed DSR.

Benchmarking Cloud Deployable DSR

This document is divided into several sections:
• Infrastructure Environment. This section provides details of the infrastructures used for the
benchmark testing, including the hardware and software. It also describes key settings and
attributes, and some recommendations on configuration.
• A benchmark section for each DSR server type. Each DSR server type is given independent
treatment for its benchmark. Each section describes the traffic setup, and the observed results. It
also provides metrics and guidelines for assessing performance on any infrastructure.

What to do with all this data?

This data is intended to provide guidance. Recommendations may need to be adapted to the conditions
in a given operator’s network. Each section below provides metrics that provide feedback on the running
performance of the application.
When planning to deploy a DSR into any cloud environment, a few steps are recommended:
• Understand the initial deployment scenario for the DSR. Which features are planned, how much of
what type of traffic? Of course, this may change once deployed, and the DSR can be grown or
shrunk to meet the changing needs.
• Use the DSR cloud dimensioning tool to get an estimate of the types of DSR virtual servers needed,
and an initial estimate of the quantity of the virtual machines and resources. Your Oracle Sales
Consultant can run this tool for you based on your DSR requirements:
• The tool allows for a very detailed model to be built of your DSR requirements including:
• Required MPS by Diameter Application ID (S6a, Sd, Gx, Rx, etc.).
• Required DSR applications such as Full Address Based Resolution (FABR) and Policy DRA
(PDRA) and any required sizing information such as the number of subscribers supported for
each application.
• Any required DSR features such as Topology Hiding, Message Copy, IPSEC, or Mediation
that can affect performance.
• Network-level redundancy requirements, such as mated pair DSR deployments where one
DSR needs to support full traffic when one of the DSRs is unavailable.
• Infrastructure information such as VMware vs. KVM, and Server parameters.
• The tool then generates a recommended number of VMs for each of the required VM types.

DSR 8.4 Cloud Benchmarking Guide 11

 • As noted below, these recommendations are just guidelines, since the actual performance of the
DSR can vary significantly based on the details of the infrastructure.
• Based on the initial deployment scenario, determine if additional benchmarking is warranted:
• For labs and trials, there is no need to benchmark performance and capacity if the goal of the lab
is to test DSR functionality.
• If the server hardware is different from the hardware used in this document then the performance
differences can likely be estimated using industry standard metrics comparing single-threaded
processor performance of the CPUs used in this document versus the CPUs used in the
customer’s infrastructure. This approach is most accurate for small differences in hardware (for
instance, different clock speeds for the same generation of Intel processors) and least accurate
across processor generations where other architectural differences such as networking interfaces
could also affect the comparison.
• It is the operator’s decision to determine if additional benchmarking in the operator’s infrastructure
is desired. Here’s a few things to consider when deciding:
• Benchmark infrastructure is similar to the operator’s infrastructure, and the operator is
satisfied with the benchmark data provided by Oracle.
• Initial turn-up of the DSR is handling a relatively small amount of traffic and the operator
prefers to measure and adjust once deployed.
• Operator is satisfied with the high-availability and geo-diversity of the DSR, and is willing to
risk initial overload conditions, and adjusts once the DSR is production.
• If desired, execute benchmarking testing on the target cloud infrastructure. Only benchmark those
types of DSR servers needed for the deployment (for example, if full address resolution is not
planned, don’t waste time benchmarking the SDS, SDS SOAM, or DPs).
• Once that benchmarking is completed, take a look at the data for each server type, and compare
it to the baseline used for the estimate (from the cloud dimensioning tool).
• If the performance estimate for a given DSR function is X and the observed performance is Y,
then adjust the performance for that DSR function to Y.
• Recalculate the resources needed for deployment based on the updated values.
• Deploy the DSR.
• Monitor the DSR performance and capacity as described later in the document. As the network
changes additional resources may be required. Increase the DSR resources as described later in this
document as needed.

Infrastructure Environment
This section describes the infrastructure that was used for benchmarking. In general, the defaults or
recommendations for hypervisor settings are available from the infrastructure vendors (for example, ESXi
vendor recommendations and defaults found in [[1]]. Whenever possible the DSR recommendations
align with vendor defaults and recommendations. Benchmarking was performed with the settings
described in this section. Operators may choose different values; better or worse performance compared
to the benchmarks may be observed. When recommendations other than vendor defaults or
recommendations are made, additional explanations are included in the applicable section.
There is a subsection included for each infrastructure environment used in benchmarking.

DSR 8.4 Cloud Benchmarking Guide 12

General Rules for All Infrastructures
Hyper-Threading and CPU Over-Subscription
All of the tests were conducted with Hyper-Threading enabled, and a 1:1 subscription ratio for vCPUs in
the hypervisor. The hardware used for the testing were dual-processor servers with 18 physical cores
each (Xeon E5-2699v3). Thus each server had:
(2 CPUs) x (18 cores per CPU) x (2 threads per core) = 72 vCPUs
It is not recommended to use over-subscribed vCPUs (for instance 4:1) in the hypervisor. Not only is the
performance lower, but it makes the performance more dependent on the other loads running on each
physical server.
Turning off Hyper-Threading is also not recommended. There ia a small increase in performance of a
given VM without Hyper-Threading for a given number of vCPUs. But since the number of vCPUs per
processor drops in half without Hyper-Threading, the overall throughput per server also drops almost by
half.
The vCPU sizing per VM is provided in section DSR VM Configurations.

Recommendation: Hyper-Threading enabled and 1:1 CPU subscription ratio.

CPU Technology
The CPUs in the servers used for the benchmarking were the Intel Xeon E5-2699v3. Servers with
different processors are going to give different results. In general there are two issues when mapping the
results of the benchmarking data in this document to other CPUs:
1. The per-thread performance of a CPU is the main attribute that determines VM performance since the
number of threads is fixed in the VM sizing as shown in section DSR VM Configurations. A good
metric for comparing the per-thread performance of different CPUs is the integer performance
measured by the SPECint2006 (CINT2006) defined by SPEC.ORG. The mapping of SPECint2006
ratios to DSR VM performance ratios isn’t exact, but it’s a good measure to determine whether a
different CPU is likely to run the VMs faster or slower than the benchmark results in this document.
Conversely CPU clock speeds are a relatively poor indicator of relative CPU performance. Within a
given Intel CPU generation (v2, v3, v4, etc.) there are other factors that affect per-thread performance
such as potential turbo speeds of the CPU vs. the cooling solution in a given server. Comparing
between Intel CPU generations there is a generation over generation improvement of CPU
throughput vs. clock speed that means that even a newer generation chip with a slower clock speed
may run a DSR VM faster.
2. The processors must have enough cores that a given VM can fit entirely into a NUMA node. Splitting
a VM across NUMA nodes greatly reduces the performance of that VM. The largest VM size (see
section DSR VM Configurations) is 12 vCPUs. Thus, the smallest processor that should be used is a
6 core processor. Using processors with more cores typically makes it easier to “pack” VMs more
efficiently into NUMA nodes, but should not affect individual VM CPU-related performance otherwise
(see the next note though).
3. One caveat about CPUs with very high core counts is that the user must be aware of potential
bottlenecks caused by many VMs contending for shared resources such as network interfaces and
ephemeral storage on the server. These tests were run on relatively large CPUs (18 physical cores
per chip), and no such bottlenecks were encountered while running strictly DSR VMs. In clouds with
VMs from other applications potentially running on the same physical server as DSR VMs, or in future
processor generations with much higher core counts, this potential contention for shared server
resources has to be watched closely.
Recommendation: The selected VM sizes should fit within a single NUMA node (for instance 6 physical
cores for the VMs that required 12 vCPUs. Check the performance of the target
CPU type against the benchmarked CPU using per-thread integer performance
metrics.

DSR 8.4 Cloud Benchmarking Guide 13

VM Packing Rate
The OCDSR doesn’t require or use CPU pinning. Thus the packing of the OCDSR VMs onto the physical
servers is under the control of OpenStack using the affinity/anti-affinity rules given in DSR VM
Configurations. Typically the VMs do not fit exactly into the number of vCPUs available in each NUMA
node, leaving some un-allocated vCPUs. The ratio of the allocated to the unallocated vCPUs is the VM
Packing Ratio. For instance, on a given server if 60 out 72 vCPUs on a server were allocated by the
OpenStack, that server would have a packing ratio of ~83%. The achieved packing in a deployment
depends on a lot of factors, including the mix of large VMs (DA-MPs, SBRs) with the smaller VMs, and
whether the OCDSR is sharing the servers with other applications that have a lot or large or small VMs.
When planning the number of physical servers required for an OCDRS a target packing ratio of 80% is a
good planning number. A packing ratio of 100% is hard to achieve, and may affect the performance
numbers shown in the benchmarks. Some amount of server capacity is necessary to run the Host OS for
the VMs, performing functions such as interrupt handling. A packing ratio of 95% or lower is desirable.

Recommendation: When planning for physical server capacity a packing ratio of 80% is a good
guideline. Packing ratios of greater than 95% might affect the benchmark numbers
since there aren’t sufficient server resources to handle the overhead of Host OSs.

Infrastructure Tuning
The following parameters should be set in the infrastructure to improve DSR VM performance. The
instructions for setting them for a given infrastructure is including the DSR Cloud Installation Guide [[3]].
• Txqueuelen: The default of 500 is too small. Recommendation is to set this parameter to 30,000.
• Tuned on the compute hosts.
• Default value of 500 is too small. Our recommendation is to set to 30000. This increases the
network throughput of a VM.
• Ring buffer increase on the physical Ethernet interfaces: The default is too small. The
recommendation is to set both receive and transmit values to 4096.
• Multiqueue: Multiqueue should be enabled on any IPFE VMs to improve performance. Already
enabled by default on ESXi, needs to be set for Openstack.
• Advanced NUMA settings (ESXi only): The SwapLoadEnabled and SwapLocalityEnabled options
should be disabled. This prevents the ESXi scheduler from moving VMs around from one NUMA
node to another trying to optimize performance. These settings aren’t appropriate for VMs that are
processing real-time loads since messages might be delayed during the move.

Recommendation: Follow instructions in the DSR Installation Guide.

KVM (QEMU)/Oracle X5-2 – Infrastructure Environment

There are a number of settings that affect performance of the hosted virtual machines. A number of tests
were performed to maximize the performance of the underlying virtual machines for the DSR application.
Host Hardware
• Oracle Server X5-2
• CPU Model: Intel® Xeon® CPU E5-2699 v3 @ 2.30GHz
• 2 CPUs
• 18 physical cores per CPU
• RAM: 128 GB
• HDD: 2.3 TB of solid state drive (SSD) storage

DSR 8.4 Cloud Benchmarking Guide 14

 • NIC:
• 4 x Intel Ethernet Controller 10-Gigabit x540-AT2
Hypervisor
• QEMU-KVM Version: QEMU 1.5.3, libvirt 1.2.8, API QEMU 1.2.8

Device Drivers
VirtIO is a virtualizing standard for network and disk device drivers where just the guest’s device driver
“knows” it is running in a virtual environment, and cooperates with the hypervisor. This enables guests to
get high performance network and disk operations, and gives most of the performance benefits of para-
virtualization.
Vhost-net provides improved network performance over Virtio-net by totally bypassing QEMU as a fast
path for interruptions. The vhost-net runs as a kernel thread and interrupts with less overhead providing
near native performance. The advantages of using the vhost-net approach are reduced copy operations,
lower latency, and lower CPU usage.

Recommendation: Vhost-net driver is recommended.

BIOS Power Settings

Typical BIOS power settings (hardware vendor dependent, see your infrastructure hardware vendor
documentation for details) provide three options for power settings:
• Power Supply Maximum: The maximum power the available PSUs can draw
• Allocated Power: The power is allocated for installed and hot pluggable components
• Peak Permitted: The maximum power the system is permitted to consume

Recommendation: Set to Allocated Power or equivalent for your Hardware vendor

Disk Image Formats

The two preferred disk image file formats available when deploying a KVM virtual machine:
• QCOW2: Disk format supported by the QEMU emulator that can expand dynamically and supports
Copy on Write.
• Raw Dump Representation: Unstructured disk image format.
QCOW2 provides a number of benefits over raw dump such as:
• Smaller file size, even on filesystems which don’t support holes (such as, sparse files)
• Copy-on-write support, where the image only represents changes made to an underlying disk image
• Snapshot support, where the image can contain multiple snapshots of the images history
The recommended Container format is “bare” – no container or metadata envelope for the disk image.
The container format string is not currently being used by OpenStack components.

Recommendation: QCOW2 (Since DSR does not involve processes which are disk I/O intensive.)
Recommendation: Bare Container format

Guest Caching Modes

The operating system maintains a page cache to improve the storage I/O performance. With the page
cache, write operations to the storage system are considered completed after the data has been copied
to the page cache. Read operations can be satisfied from the page cache if the data requested is in the
cache. The page cache is copied to permanent storage using fsync. Direct I/O requests bypass the page

DSR 8.4 Cloud Benchmarking Guide 15

cache. In the KVM environment, both the host and guest operating systems can maintain their own page
caches, resulting in two copies of data in memory.
The following caching modes are supported for KVM guests:
• Writethrough: I/O from the guest is cached on the host but written through to the physical medium.
This mode is slower and prone to scaling problems. Best used for a small number of guests with
lower I/O requirements. Suggested for guests that do not support a writeback cache (such as, Red
Hat Enterprise Linux 5.5 and earlier), where migration is not needed.
• Writeback: With caching set to writeback mode, both the host page cache and the disk write cache
are enabled for the guest. Because of this, the I/O performance for applications running in the guest
is good, but the data is not protected in a power failure. As a result, this caching mode is
recommended only for temporary data where potential data loss is not a concern.
• None [Selected]: With caching mode set to none, the host page cache is disabled, but the disk write
cache is enabled for the guest. In this mode, the write performance in the guest is optimal because
write operations bypass the host page cache and go directly to the disk write cache. If the disk write
cache is battery-backed, or if the applications or storage stack in the guest transfer data properly
(either through fsync operations or file system barriers), then data integrity can be ensured. However,
because the host page cache is disabled, the read performance in the guest would not be as good as
in the modes where the host page cache is enabled, such as write through mode.
• Unsafe: The host may cache all disk I/O, and sync requests from guest are ignored.
Caching mode None is recommended for remote NFS storage, because direct I/O operations
(O_DIRECT) perform better than synchronous I/O operations (with O_SYNC). Caching mode None
effectively turns all guest I/O operations into direct I/O operations on the host, which is the NFS client in
this environment. Moreover, it is the only option to support migration.
Recommendation: Caching Mode = None

Memory Tuning Parameters

Swappiness
The swappiness parameter controls the tendency of the kernel to move processes out of physical
memory and onto the swap disk. Because disks are much slower than RAM, this can lead to slower
response times for system and applications if processes are too aggressively moved out of memory.
• vm.swappiness = 0: The kernel swaps only to avoid an out of memory condition.
• vm.swappiness = 1: Kernel version 3.5 and over, as well as kernel version 2.6.32-303 and over;
Minimum amount of swapping without disabling it entirely.
• vm.swappiness = 10: This value is recommended to improve performance when sufficient memory
exists in a system.
• vm.swappiness = 60: Default
• vm.swappiness = 100: The kernel swaps aggressively.

Recommendation: vm.swappiness = 10
Kernel Same Page Merging
Kernel Same-page Merging (KSM), used by the KVM hypervisor, allows KVM guests to share identical
memory pages. These shared pages are usually common libraries or other identical, high-use data.
KSM allows for greater guest density of identical or similar guest operating systems by avoiding memory
duplication. KSM enables the kernel to examine two or more already running programs and compare
their memory. If any memory regions or pages are identical, KSM reduces multiple identical memory
pages to a single page. This page is then marked copy on write. If the contents of the page is modified
by a guest virtual machine, a new page is created for that guest.

DSR 8.4 Cloud Benchmarking Guide 16

This is useful for virtualization with KVM. When a guest virtual machine is started, it only inherits the
memory from the host qemu-kvm process. Once the guest is running, the contents of the guest operating
system image can be shared when guests are running the same operating system or applications. KSM
allows KVM to request that these identical guest memory regions be shared.
KSM provides enhanced memory speed and utilization. With KSM, common process data is stored in
cache or in main memory. This reduces cache misses for the KVM guests, which can improve
performance for some applications and operating systems. Secondly, sharing memory reduces the
overall memory usage of guests, which allows for higher densities and greater utilization of resources.
The following 2 Services control KSM:
• KSM Service: When the KSM service is started, KSM shares up to half of the host system's main
memory. Start the KSM service to enable KSM to share more memory.
• KSM Tuning Service: The ksmtuned service loops and adjusts KSM. The ksmtuned service is
notified by libvirt when a guest virtual machine is created or destroyed.

Recommendation: KSM service set to active and ksmtuned service running on KVM hosts.
Zone Reclaim Mode
When an operating system allocates memory to a NUMA node, but the NUMA node is full, the operating
system reclaims memory for the local NUMA node rather than immediately allocating the memory to a
remote NUMA node. The performance benefit of allocating memory to the local node outweighs the
performance drawback of reclaiming the memory. However, in some situations reclaiming memory
decreases performance to the extent that the opposite is true. In other words, in these situations,
allocating memory to a remote NUMA node generates better performance than reclaiming memory for the
local node.
A guest operating system causes zone reclaim in the following situations:
• When you configure the guest operating system to use huge pages.
• When you use KSM to share memory pages between guest operating systems.
Configuring huge pages and running KSM are both best practices for KVM environments. Therefore, to
optimize performance in KVM environments, it is recommended to disable zone reclaim.
Recommendation: Disable Zone Reclaim.

Transparent Huge Pages

Transparent huge pages (THP) automatically optimize system settings for performance. By allowing all
free memory to be used as cache, performance is increased.
Recommendation: Enable THP.

VMware (ESXi) – Infrastructure Environment

There are a number of ESXi (VMware hypervisor) settings that affect performance of the hosted virtual
machines. A number of tests were performed to maximize the performance of the underlying virtual
machines for the DSR application.
• Hypervisor versions tested
• ESXi 5.5
• ESXi 6.0 U2

DSR 8.4 Cloud Benchmarking Guide 17

Virtual Sockets vs. Cores per Virtual Socket
When defining a virtual machine the number of vCPUs must be assigned to a server. The user has
options for setting the number of “Virtual Sockets” and the number of “Cores per Virtual Socket”. The
product of these two parameters determines the number of vCPUs available to the virtual machine.
In following the VMware best practices, the default value of 1 core per socket was used. This
configuration is referred to as “wide” and “flat.” This enables vNUMA to select and present the best virtual
NUMA topology to the guest operating system, which is optimal on the underlying physical topology.
Recommendation: 1 core per socket, virtual socket set to the number of vCPUs required by the server
role.

Network Settings
Network Adapters
There is a number of networking adapter choices when deploying a virtual machine:
• E1000: This adapter is an emulated version of Intel 82545EM Gigabit Ethernet Controller.
VMXNET3 adapter is the next generation of Para virtualized NIC designed for performance.
• VMXNET3: This adapter has less CPU overhead compared to e1000 or e1000e. Also, VMXNET3 is
more stable than e1000 or e1000e. VMXNET3 adapter is the next generation of Para virtualized NIC
designed for performance. This is the vSphere default setting.
• VMXNET family implements an idealized network interface that passes network traffic between the
virtual machine and the physical network interface cards with minimal overhead.
Recommendation: VMXNET3. No observable differences were noticed between E1000 and VMXNET3
for DSR application testing.
Virtual Network Interrupt Coalescing and SplitRx Mode
• Virtual network Interrupt Coalescing: This option reduces number of interrupts thus potentially
decreasing CPU utilization. This may however increase network latency. By default this is enabled in
ESX 5.5 and 6.0.
• SplitRxMode: This option uses multiple physical CPUs to process network packets received in single
network queue. By default this is enabled in ESX 5.5 and 6.0 for VMXNET3 adapter type.
Table 3: Virtual Network Interrupt Coalescing and SplitRX Mode
Virtual Network Interrupt SplitRxMode:
Network Setting Default Coalescing: Disabled Disabled
DSR.CPU (Avg/Max) ~40.7%/~44.5% ~42%/~45.5% ~38.8%/~40.6%
System.CPU_UtilPct (Avg/Max) ~44.4%/~53% ~44.4%/~55.5% ~41.8%/~53.3%
Latency Observed as same in DSR application benchmarking

Recommendation: Virtual network interrupt coalescing: Enabled; SplitRxMode: Enabled.

DSR 8.4 Cloud Benchmarking Guide 18

Power Settings
VMware ESXi allows assignment of power management profiles. These profiles allow the user to
configure the host to save power while balancing performance. The power management profiles use the
host’s processor ACPI power setting. Many host manufacturer’s bios overrides the ESXi settings.
Table 4: Power Management Profiles
ESXi Power Mode High Performance Balanced Performance
System.CPU UtilPct (Avg/Max) ~40%/~60% ~38%/~55%
Dsr.Cpu (Avg/Max) ~38%/~48% ~36%/~44%
Used %/Run %/System % ~472/~388/~49% ~462/~376/~49%
Wait %/Idle % ~407%/~1013 ~419%/~1023

The data in the table above is collected from a DA MP, but similar trends are observed on the other DSR
virtual server types. A small but significant difference was observed between balanced and high
performance power settings. However, the data did not indicate a large enough deviation to vary from the
hardware vendor’s guidelines. DSR benchmark testing was performed with ESXI Power Mode set to
Balanced Performance.

Recommendation: Refer to host hardware vendor power management guidelines for virtualization.

Hardware Assisted Virtualization

VMware ESXi automatically determines if a virtual machine can use hardware support for virtualization
based on processor type. Several settings were selected for assessing performance:
A. Automatic
B. Use software for instruction set and MMU virtualization.
C. Use Intel® VT-x/AMD-V™ for instruction set virtualization and software for MMU virtualization
D. Use Intel® VT-x/AMD-V™ for instruction set virtualization and Intel® EPT/AMD RVI for MMU
virtualization
Also testing with “Node Interleaving” setting Enabled (that is, NUMA disabled), with no noticeable
changes in performance.
Table 5: Virtualization Performance by Processor
MMU Virtualization Setting A. B. C. D.
System CPU UtilPct (Max/Avg) 57.5/38% 71.5/43% 71.5/43% 53/38%
Dsr.Cpu (Max/Avg) 43.5/36.3% 50/38.6% 50/38.5% 43/36.3%
The data in the table above is provided from a DA MP. Trends for other servers are similar. The
automatic (default) settings provide performance better than options B and C above, and fairly equivalent
to option D.
Recommendation: Refer to host hardware vendor guidelines for virtualization. Defaults recommended.

Virtual Machine Storage Configuration

Storage Type Adapter
Testing was performed with the default “LSI Logic Parallel” option. No testing was performed against
recent virtual SCSI adapters (LSI Logic SAS and VMware para-virtualized (PVSCSI.) At the time of
testing the default was considered as the most stable and compatible.

Recommendation: Default “LSI Logic Parallel”

DSR 8.4 Cloud Benchmarking Guide 19

Disk Provisioning
The following disk provisioning options are available when deploying a virtual machine:
• Thick Provision Lazy Zeroed: All space needed for the VM is allocated during creation. Data on the
host disk is zeroed out at a later time on first write from the virtual machine.
• Thick Provision Eager Zeroed: All space needed for the VM is allocated during creation. The data on
the host disk is zeroed out during creation. Time to deploy the virtual machine id increased with this
option. This option supports fault tolerant features provided by the infrastructure.
• Thin Provision: This option uses only the amount needed by the virtual machine disk. The image
grows as needed until the allocated capacity is reached.
With the high availability of the DSR, storage should be allocated at the time the VM is created, so thin
provisioned is not recommended. When instantiating a fairly typical DSR VM with 60G of storage, the
lazy zeroed disk was created almost instantaneously. Whereas the eager zeroed disk took about 7
minutes to initialize. Lazy zeroed is recommended.

Recommendation: Thick Provisioned Lazy Zeroed.

Large Memory Pages

VMware ESXi Large-page support enables server applications to establish large-page memory regions.
The use of large pages can potentially increase TLB access efficiency and thus improve program
performance. By default Large page support is enabled on VMware ESXi Server although the full benefit
of large pages comes only when guest OS and applications use them as well. By default large page
support is not enabled in the DSR product.
The following settings were evaluated:
• Default settings (such as, Large memory pages support enabled on host and large pages configured
as 0 on guest).
• Large memory pages support enabled on host and 1024 large pages configured on guest.
• Large memory pages support disabled on host.

Recommendation: Default settings. No visible advantage was observed when comparing iterative
memory stats as observed through /proc/meminfo. No visible advantage could be
observed in using large pages.

Benchmark Testing
The way the testing was performed and the benchmark test set-up is the same for each benchmark
infrastructure. Each section below describes the common set-up and procedures used to benchmark,
and then the specific results for the benchmarks are provided for each benchmark infrastructure. In
general the benchmarking results for VMware/ESXi vs. Openstack/KVM are close enough that only one
set of numbers are shown.

DA MP Relay Benchmark
Overview
This benchmarking case illustrates conditions for an overload of a DSR DA MP. Simulator message rate
is increased until the DA-MP Overload mechanisms are triggered causing messages to be discarded.

DSR 8.4 Cloud Benchmarking Guide 20

Topology

DA-MP-01
IPFE-01

...
MME Simulator HSS Simulator
IPFE-02
DA-MP-X

Figure 1: DA-MP Testing Topology

Figure 1 illustrates the logical topology used for this testing. Diameter traffic is generated by an MME
simulator and sent to an HSS simulator. The traffic is ramped up in increments until congestion is seen
on the DA-MPs indicating that the maximum normal traffic handling capacity has been reached for the
current message flow.

The dsr.cpu utilization can be increased from dsr.cpu 53% to higher levels by means of configuration
changes with DOC/CL1/CL2 discards set to 0 and multi queuing enabled on all hosts. With this
configuration, it must be noted that all the discards will be at one step CL3 for all incoming and outgoing
messages. The Default CPU threshold configuration shall remain the same which is DOC: 54% CPU ;
CL1: 60% CPU ; CL2: 66% CPU.
The Relay traffic with the above configuration changes is measured at 36K MPS at 66% dsr.cpu for a DA-
MP profile as defined in Appendix A. DSR VM Configuration. The hyper threading was enabled

Message Flow
Figure 2 illustrates the Message sequence for this benchmark case.

MME DSR HSS

ULR

ULR

ULA

ULA

Figure 2: DA-MP Message Sequence

Table 6: DA-MP Test Set Up

Messages Traffic Details

Message Distribution Detail Distribution
ULR, ULA 100% Average Message Size 2.2K
Cross DSR Routing 75% or higher

Indicative Alarms/Events
During benchmark testing the following alarms/events were observed as it crossed into congestion.
Table 7: DA-MP Alarms/Events
Number Severity Server Name Description

DSR 8.4 Cloud Benchmarking Guide 21

 5005 Minor IPFE IPFE Backend in A backend server not accepting new
Stasis connections but continuing to process
existing connections
22200 Major DA-MP Communication Agent The Diameter process is approaching or
Routed Service exceeding its engineered traffic handling
Congested capacity
22215 Major DA-MP Ingress Message Ingress message discarded due to DA MP
Discarded: DA MP (danger of) CPU congestion.
Overload Control

DA-MP Indicators for Capacity Expansion

The following DSR system measurements, KPIs and events can be used to monitor the performance of
the DA-MP in a field deployment, or to drive the orchestration of DA-MP expansion.

DA-MP Utilization
In this section, only the key recommended metrics for planning expansions of the DA-MP are discussed.
There are many more measurements available on the DA-MP, and these can be found in [[2]].
The key metrics for managing the DA-MP are:
Table 8: DA-MP Utilization Metrics
Recommended Usage
Measure-
ment ID Name Group Scope Description Condition Actions
10204 EvDiameterProcessAvg MP Server Average When running in If additional growth in the system
Performance Group percent normal operation with a is anticipated, then consider
Diameter mate in normal adding an additional DA-MP.
Process CPU operation, and this It is possible that the traffic mix is
utilization (0- measurement exceeds different than originally
100%) on a MP 30% of the rated dimensioned (for example, 40%
server maximum capacity, IPSEC instead of the originally
OR dimensioning 5%). In these cases,
Exceeds 60% of the re-assess the dimensioning with
rated capacity when the actual traffic/application mix
running without an and add additional DA-MPs as
active mate. needed.
10205 TmMpCongestion MP Server Total time (in Any number greater After eliminating any configuration,
Performance Group milliseconds) than 0 anomalous traffic spikes or major
spent in local failure conditions, then is a late
MP congestion indication that additional DA MPs
state are needed.
It is highly desirable that planning
for additional DA-MPs happens
before this condition occurs.
10133 RxMsgSizeAvg Diameter Server The average Average message size DA-MP dimensioning assumes 2K
Performance Group ingress > 2000 bytes average message size. This
message size information is used to dimension
in Diameter IPFEs and DIH/IDIH. No action
payload octets required if there are no alarms
associated with the PDU message
pool (available memory for
messages). If PDU message pool
is exhausting, contact Oracle.
31056 RAM_UtilPct_Average System System The average If the average Ram Contact Oracle
committed utilization exceeds
RAM usage as 80% utilization
a percentage of
the total
physical RAM

DSR 8.4 Cloud Benchmarking Guide 22

 Recommended Usage
Measure-
ment ID Name Group Scope Description Condition Actions
31052 CPU_UtilPct_Average System System The average When running in If additional growth in the system
CPU usage normal operation with a is anticipated, then consider
from 0 to 100% mate in normal adding an additional DA MP.
(100% operation, and this It’s possible that the traffic mix is
indicates that measurements different than originally
all cores are exceeds 30% of the dimensioned (for example, 40%
completely rated maximum IPSEC instead of the originally
busy) capacity, dimensioning 5%). In these cases,
OR re-assess the dimensioning with
Exceeds 60% of the the actual traffic and application
rated capacity when mix and add additional DA-MPs
running without an VMs as needed.
active mate.

Measuring DA-MP Connection Utilization

In this section, only the key recommended metrics for planning expansions of the DA-MP connections are
discussed. There are many more measurements available on the DA-MP connections, and these can be
found in [[2]].
The key metrics for managing the DA-MP connections are:
Table 9: DA-MP Connection Metrics

Measure- Recommended Usage

ment ID Name Group Scope Description Condition Actions
10500 RxConnAvgMPS Connection Connection Average Minor alarm is set by Configure
Performance Ingress default at 50%, major at additional
Message Rate 80%. Ingress message connections
(messages rate per connection is
per second) customer configurable with
utilization on a a max per connection of
connection 10,000

Suggested Resolution
If congestion alarms shown in Table 7: DA-MP Alarms/Events, then add additional DA-MPs to avoid CPU
congestion. However, if the connection alarm shown in Table 9: DA-MP Connection Metrics is seen,
then adding additional connections for that peer helps distribute the load and alleviates the connection
alarm.
In general, the growth mechanism for DA MPs is via horizontal scaling. That is by adding additional DA
MPs. The current maximum number of the DA MPs per DSR signaling NE is 32.

Full Address Based Resolution (FABR) Capacity

Overview
The FABR application adds a Database Processor (DP) server is to perform database lookups with a user
defined key (IMSI, MSISDN, or Account ID and MSISDN or IMSI.) If the key is contained in the database,
the DP returns the realm and FQDN associated with that key. The returned realm and FQDN can be
used by the DSR Routing layer to route the connection to the desired endpoint. Since there is additional
work done on the DA-MP to query the DP, running the FABR application has an impact on the DA-MP
performance. This section contains the performance of the DA-MP while running FABR as well as
benchmark measurements on the DP itself.

DSR 8.4 Cloud Benchmarking Guide 23

Topology

SDS

DP

IPFE-01
MME
DA-MPs HSS Simulators
Simulators
IPFE-02

Figure 3: SDS DP Testing Topology

SDS DB Details
The SDS database was first populated with subscribers. This population simulates real-world scenarios
likely encountered in a production environment and ensure the database is of substantial size to be
queried against.
• SDS DB Size: 300 Million Routing Entities (150 M MSISDNs/150 M IMSIs)
• AVP Decoded: User-Name for IMSI
New SDS profile (Large) is introduced to enhance the capacity of SDS FABR database to 1 Billion
Routing Entities. The Large profile is defined in Appendix A. DSR VM Configurations based on the below
1 Billion Entry configuration:
• 260 Million Subscribers having 2 IMSI, 1 MSISDN = 780 Million Routing entities. IMSI = 15 bytes,
MSISDB= 11 bytes.
• 220 Million IOT records, of 27 bytes each.
• Destinations: 300 Entries.
• One Destination per Routing Entity is configured.
• Longest FQDN configured: 32 characters
• Longest Realm configured: 13 characters

Message Flow

MME DSR SDS HSS

ULR (S6a)

Sub scribe r Qu ery Requ est

Sub scribe r Qu ery Response

ULR (S6a)

ULA (S6a)

ULA (S6a)

Figure 4: SDS DP Message Sequence

Table 10: SDS DP Message Details

DSR 8.4 Cloud Benchmarking Guide 24

 Messages Traffic Details
Message Distribution Detail Distribution
ULR, ULA 100% Average Message Size 2.2K
Cross DSR Routing 75% or higher

DP Indicators for Capacity Expansion

Table 11: SDS DP Alarms/Events
Number Severity Server Name Description
19900 Minor sdsDP Process CPU Utilization The Process, which is responsible
for handling all traffic, is
approaching or exceeding its
engineered traffic handling
capacity.
19822 Major DA-MP Communication Agent Communication Agent Routed
Routed Service Congested Service Congested.
19825 Major/Critical DA-MP Communication Agent The number of failed transactions
Transaction Failure Rate during the sampling period has
exceeded configured thresholds.
19826 Major sdsDP Communication Agent Communication Agent Connection
Connection Congested Congested.
19831 Info DA-MP Communication Agent Communication Agent Service
Service Operational State Operational State Changed,
Changed Instance DPService.
19816 Info DA-MP Communication Agent Configuration Mode = Configured,
Connection state Changed Admin State = Enabled, Connect
Mode = Server, Operational State =
Degraded, Congestion Level = 1,
Overload Level = 1, Transport
Congestion Level = 0.

Measuring DP Utilization
In this section, only the key recommended metrics for managing the performance of the DP are
discussed. There are many more measurements available on the DP, and these can be found in [[2]].
There are two key components of the subscriber database within a DSR Signaling node: the Database
Processors (DPs), and OAM component which runs on the System OAM VMs. The key metrics for
managing the DPs are:
Table 12: SDS DP Utilization Metrics
Recommended Usage
Measure-
ment ID Name Group Scope Description Condition Actions
4170 DpQueriesReceived DP System The total When running in normal The operator should
(per number of operation with a mate in normal determine if additional growth
DP) queries received operation, and this in the number traffic requiring
per second measurement exceeds 30% of subscriber database look-ups
the benchmarked maximum is continuing to grow. If so,
capacity, an estimate of the additional
OR rate of database lookups
Exceeds 60% of the should be calculated and
benchmarked capacity when additional DPs should be
running without an active mate. planned for.

DSR 8.4 Cloud Benchmarking Guide 25

 Recommended Usage
Measure-
ment ID Name Group Scope Description Condition Actions
31056 RAM_UtilPct_Average System System The average If the average Ram utilization Contact Oracle
(per committed RAM exceeds 80% utilization
DP) usage as a
percentage of
the total physical
RAM
31052 CPU_UtilPct_Average System System The average When running in normal Oracle considers this
(per CPU usage from operation with a mate in normal measurement of lesser
DP) 0 to 100% operation, and this importance to the
(100% indicates measurements exceeds 30% DpQueriesReceived.
that all cores are of the rated maximum capacity, However, this measurement
completely busy) OR in conjunction with
DpQueriesReceived can be
Exceeds 60% of the
benchmarked capacity when used to indicate the need to
running without an active mate. add additional DPs.

While memory is a consideration for the DPs, the SDS provides the centralized provisioning for the entire
DSR network.
The OAM application related to the DPs (DP SOAM) runs at each DSR Signaling NE requiring the Full
Address Resolution feature. Currently these are fixed sized VMs with no horizontal or vertical scaling
recommended as no need for scaling these VMs has been observed. The following two metrics should
be monitored,
Table 13: DP SOAM Metrics

Measure- Recommended Usage

ment ID Name Group Scope Description Condition Actions
31056 RAM_UtilPct_Average System System The average If the average Contact
(per DP committed RAM usage Ram utilization Oracle
SOAM) as a percentage of the exceeds 80%
total physical RAM utilization
31052 CPU_UtilPct_Average System System The average CPU If the average Contact
(per DP usage from 0 to 100% CPU utilization Oracle
SOAM) (100% indicates that exceeds 80%
all cores are utilization
completely busy)

Suggested Resolution
The growth mechanism for DPs is via horizontal scaling by adding additional DPs. The current maximum
number of the DPs per DSR signaling NE is 10. This amount of scaling currently well exceeds capacities
of the DA MPs driving queries to the DPs.

SS7 MP
Overview
SS7-MP is not supported since DSR 8.3. The SS7-MP server type is responsible for transforming
messages between SS7 and Diameter protocols. Both Diameter and MAP messages were sent from the
simulator to the DSR. The SS7-MP is based on similar technology to the DA-MP benchmarked in
previous sessions. The licensed capacity of the SS7-MP is currently limited to 12k MPS per SS7-MP,
even though the performance is similar to the DA-MP. This section is here to explain the functions of the
SS7-MP, and note the SS7-MP specific events and measurements used to monitor it.

DSR 8.4 Cloud Benchmarking Guide 26

Topology

IPFE-01
MME/SGSN
DA-MPs SS7-MPs* HLR Simulator
Simulator
IPFE-02

*System under test

Figure 5: SS7 MP Testing Topology

Message Flow
MME /
DSR SS7-MP HLR
SGSN

ULR (S6a)

IWF Re quest w/ULR

UpdateGp rsL oca tion Arg

(translated)

UpdateGp rsL oca tion Re s

ULA (translated)

ULA (S6a)

Purge_MSArg

PUR (translated)

PUR (S6a)

PUA (s6a)

IWF Response w/PUA

Purge_MSReg (translated)

Figure 6: SS7 MP Message Flow

Table 14: SS7 MP Message Detail

Detail Distribution
Diameter originated 50%
MAP originated 50%

Indicative Alarms/Events
Table 15: SS7 MP Alarms/Events
Number Severity Server Name Description
19250 Minor DA-SS7MP SS7 Process The SS7 Process, which is responsible for
CPU Utilization handling all SS7 traffic, is approaching or
exceeding its engineered traffic handling
capacity.

DSR 8.4 Cloud Benchmarking Guide 27

 SS7 MP Indicators for Capacity Expansion
Measuring SS7 MP Utilization
In this section, only the key recommended metrics for planning expansions of the SS7 MP are discussed.
There are many more measurements available on the SS7 MP, and these can be found in [[2]].
The key metrics for managing the SS7 MP and associated IWF DA-MP are:
Table 16: SS7 MP Metrics
Measure- Recommended Usage
ment ID (or
KPI ID) Name Group Scope Description Condition Actions
11054 MAP Ingress Message MAP SS7 Average number When running in If additional growth in MAP
Rate Diameter MP of MAP normal operation with a traffic is expected, an estimate
Interworking messages (both mate in normal of the traffic should be
function requests and operation, and this calculated and additional SS7
(MD-IWF) responses) measurement exceeds MPs should be planned for.
received per 30% of benchmarked This condition could also be an
second from maximum, anomalous spike in traffic, and
SS7 network OR exceeds 60% of the the operator may choose to
benchmarked capacity ignore the occurrence.
when running without Above 40% in normal operation
an active mate. indicates an immediate need for
additional SS7 MPs
31056 RAM_UtilPct_Average System System The average If the average Ram Contact Oracle
(SS7 MP) committed RAM utilization exceeds 80%
usage as a utilization
percentage of
the total physical
RAM
31052 CPU_UtilPct_Average System System The average When running in If additional growth in the
(SS7 MP) CPU usage from normal operation with a system is anticipated, then
0 to 100% mate in normal consider adding an additional
(100% indicates operation, and this SS7 MP.
that all cores are measurements
completely exceeds 30%,
busy) OR
Exceeds 60% when
running without an
active mate.

Suggested Resolution
Add additional SS7 MPs to accommodate additional MAP traffic.
The growth mechanism for SS7 MPs is via horizontal scaling by adding additional SS7 MPs. There can
be up to 8 SS7 MPs in a DSR.

vSTP MP
Overview
The vSTP-MP server type is a virtualized STP that supports M2PA, M3UA, EIR. It can be deployed either
with other DSR functionality as a combined DSR/vSTP, or as a standalone virtualized STP without any
DSR functionality.
The vSTP MP requires 8 GB of RAM. Up to 32 vSTP MPs can be configured, with a capacity of 10k MPS
per vSTP MP giving a total MPS of 300k. For EIR application, the capacity of 2k MPS per vSTP MP
giving a total MPS of 64k.

DSR 8.4 Cloud Benchmarking Guide 28

 vSTP MP Indicators for Capacity Expansion
Utilization
In this section, only the key recommended metrics for planning expansions of the VSTP-MP are
discussed. There are many more measurements available on the VSTP-MP, and these can be found in
[[2]].
The key metrics for managing the VSTP-MP are:
Table 17: VSTP-MP Utilization Metrics
Recommended Usage
Measure-
ment ID Name Group Scope Description Condition Actions
21150 VstpANSIPDUUtilAvg VSTP Site The average SS7 A PDU is allocated to each If both the peak and average
Server ANSI PDU Buffer message that arrives at a measurements for multiple
Resource Pool utilization (0- vSTP and is de-allocated vSTPs within a Network Element
100%) measured when message processing are consistently near the
during the completes. This recommended maximum
collection interval measurement is useful for engineered capacity of a vSTP
evaluating whether when the ingress Message Rate
21151 VstpITUPDUUtilAvg VSTP Site The average SS7
persistent network problems and/or SS7 Process CPU
Server ITU PDU Buffer
exist. In general, PDU Utilization measurement are
Resource Pool utilization (0-
buffers are engineered to below the recommended
100%) measured
match the processing engineered capacity of a vSTP,
during the
capacity of the vSTP. If then a network (IP or SS7)
collection interval
network problems exist, problem may exist. Looking at
egress messages from the these measurements on a time of
vSTP are delayed, then day basis may provide additional
PDUs/messages also sit in insight into potential network
internal SS7 queues. problems.
31056 RAM_UtilPct_Average System System The average If the average Ram Contact Oracle.
(vSTP committed RAM utilization exceeds 80%
MP) usage as a utilization
percentage of the
total physical RAM
31052 CPU_UtilPct_Average System System The average CPU When running in normal If additional growth in the system
(vSTP usage from 0 to operation with a mate in is anticipated, then consider
MP) 100% (100% normal operation, and this adding an additional vSTP MP.
indicates that all measurements exceeds It’s possible that the traffic mix is
cores are 30% of the rated maximum different than originally
completely busy) capacity, dimensioned. In these cases, re-
OR assess the dimensioning with the
Exceeds 60% of the rated actual traffic and application mix
capacity when running and add additional VSTP-MPs
without an active mate. VMs as needed.

Suggested Resolution
In general, the growth mechanism for vSTP MPs is via horizontal scaling, that is by adding additional
vSTP MPs. The current maximum number of vSTP MPs per DSR signaling NE is 32.

Policy DRA (PDRA) Benchmarking

Overview
The Policy DRA (PDRA) application adds two additional database components, the SBR(session) (SBR-
s) and the SBR(binding) (SBR-b). The DA-MP performance was also measured since the PDRA
application puts a different load on the DA-MP than either running Relay or FABR traffic. There are two
sizing metrics when determining how many SBR-s or SBR-g server groups (for example, horizontal
scaling units) are required. The first is the MPS traffic rate seen at the DA-MPs in Figure 7. This is the
metric that is benchmarked in this document. The second factor is the number of bindings (SBR-b) or

DSR 8.4 Cloud Benchmarking Guide 29

sessions (SBR-s) that can be supported. This session/binding capacity is set primarily by the memory
sizing of the VM, and is fixed at a maximum of 16 million per SBR from the DSR 8.3 release. The number
of bindings and sessions required for a given network are customer dependent. But a good starting place
for engineering is to assume:
• The number of bindings is equal to the number of subscribers supported by the PCRFs.
• The number of sessions is equal to number of subscribers times the number of IPCAN sessions
required on average for each subscriber. For instance, a subscriber might have one IPCAN session
for LTE, and one for VoLTE. Note the number of sessions is always be equal to or greater than the
number of bindings.

Topology
SBR-s

PCEF Simulators IPFE-01

DA-MPs PCRF Simulators
AF Simulators IPFE-02

SBR-b

Figure 7: SBR Testing Topology

DSR 8.4 Cloud Benchmarking Guide 30

Message Flow

CTF AF DSR OCS

CCR-I

CCR-I

CCA-I

CCA-I

CCR-U

CCR-U x4

CCA-U

CCA-U

RAR

RAR x2

RAA

RAA

CCR-T

CCR-T

CCA-T

CCA-T

AAR

AAR

AAA

AAA

RAR

RAR x2

RAA

RAA

STR

STR

STA

STA

CCR-T

CCR-T

CCA-T

CCA-T

Figure 8: PDRA Message Sequence

Table 18 shows the call model used for the testing. The message distribution is Oracle’s baseline
benchmarking, and may differ significantly from customer distributions based on factors such as the
penetration of LTE support vs. VoLTE support. The Traffic Details shows the configured PDRA options.
For more details on these options please see [[4]].

DSR 8.4 Cloud Benchmarking Guide 31

 Table 18: PDRA Test Call Model

Messages Traffic Details

Message Distribution Message Distribution
CCR-I, CCA-I 8.5% Gx w/ IPv6 Alternate Key 100%
CCR-U, CCA-U 25.5% Gx w/ IPv4 Alternate Key 0%
CCR-T, CCA-T 8.5% Gx with MSISDN Alternative Key 100%
Gx RAR, RAA 25.5% Gx Topology Hiding 0%
AAR, AAA Initial 12.8% Rx Topology Hiding 0%
STR, STA 12.8%
Rx RAR, RAA 6.4%

Indicative Alarms/Events
Table 19: SBR (b) Alarms/Events
Number Severity Server Name Description
19825 Minor/Major/Critical DA-MP Communication Agent The number of failed transactions
Transaction Failure during the sampling period has
Rate exceeded configured thresholds.
19826 Major DA-MP, Communication Agent Communication Agent
SBR(s) Connection Congested Connection Congested
19846 Major DA-MP, Communication Agent Communication Agent Resource
SBR(s) Resource Degraded Degraded
22051 Critical SOAM Peer Unavailable Unable to access the Diameter
Peer because all of the diameter
connections are Down.
22101 Major SOAM Connection Unavailable Connection is unavailable for
Diameter Request/Answer
exchange with peer.
22715 Minor SBR(s) SBR Audit Suspended SBR audit is suspended due to
congestion.
22725 Minor/Major SBR(s) SBR Server In SBR server operating in
Congestion congestion.
22732 Minor/Major SBR(s) SBR Process CPU SBR process CPU utilization
Utilization Threshold threshold has been exceeded.
Exceeded

DSR 8.4 Cloud Benchmarking Guide 32

 Measurements
Key metrics for managing the Session SBR(b) VMs are:
Table 20: Session SBR (b) VMs Metrics
Recommended Usage
Measure-
ment ID Name Group Scope Description Condition Actions
31052 CPU_UtilPct_Average System System The average CPU When this measurement exceeds 60% Contact
(SBR) usage from 0 to 100% utilization Oracle
(100% indicates that
all cores are
completely busy).
31056 RAM_UtilPct_Average System SBR The average If the average Ram utilization exceeds Contact
committed RAM 80% utilization Oracle
usage as a
percentage of the
total physical RAM.
11372 SbrPolicySessionRecsAvg SBR Session Server The number of policy If PDRA function is enabled and OC- Contact
Performance Group sessions in progress DRA is not enabled and average Oracle
exceeds benchmarked capacity.
If both PDRA and OC-DRA are enabled
this average must be combined with the
SbrOcSessionRecsAvg and the
combined average exceeds
benchmarked capacity.
11441 SbrOcSessionRecsAvg SBR Session Server The number of online If OC-DRA function is enabled and Contact
Performance Group Charging sessions in PDRA is not enabled and average Oracle
progress exceeds benchmarked capacity.
If both PDRA and OC-DRA are enabled
this average must be combined with the
SbrPolicySessionRecsAvg and the
combined average exceeds
benchmarked capacity.

Key metrics for managing the Binding SBR(b) servers are:

Table 21: Binding SBR (b) Server Metrics
Recommended Usage
Measure-
ment ID Name Group Scope Description Condition Actions
31052 CPU_UtilPct_Average System System The average CPU usage When this Contact
(blade) from 0 to 100% (100% measurement Oracle
indicates that all cores are exceeds 60%
completely busy) occupancy.
31056 RAM_UtilPct_Average System Blade The average committed RAM If the average Ram Contact
usage as a percentage of the utilization exceeds Oracle
total physical RAM 80% utilization
11374 SbrPolicyBindingRecsAvg SBR Binding Server Average number of active When this average Contact
Performance Group SBR Policy bindings exceeds Oracle
benchmarked
capacity.

Suggested Resolution
If either additional Bindings or MPS capacity is required is required then additional Server Groups may be
added to an existing SBR(b) using the SBR reconfiguration feature. There can be up to 8 Server Groups
in the SBR(b).

DSR 8.4 Cloud Benchmarking Guide 33

 Diameter Security Application (DSA) Benchmarking
Diameter Security application (DSA) applies countermeasures for ingress messages received from
external foreign network and for egress messages sent to external foreign network. Different
countermeasure profiles can be created for different IPX or roaming partners by enabling/disabling
countermeasures individually for different IPX provider or roaming partner Diameter Peers. DSA
application is enabled on DA-MP and it uses SBR-u to store context information.

Topology

Figure 9: DSA Testing Topology

The following stateful and stateless counter measure application configuration and the modes of
operations used in benchmarking tests.
Table 22: Stateful and Statelss Counter Measures
Application Configuration Data General Options Settings
Count of
Table Name Configured Entries Options Values
AppCmdCst_Config 2 Opcodes Accounting Disabled
AppIdWL_Config 1 Max. U-SBR Queries per Message 10
AVPInstChk_Config 48 Max. Size of Application State 1600
Foreign_WL_Peers_Cfg_Sets 19 Logging of Vulnerable Messages Enabled
MCC_MNC_List 139
MsgRateMon_Config 8
Realm_List 112
Security_Countermeasure_Config 15
SpecAVPScr_Config 60 Application Threads
System_Config_Options 1 Request 6
TimeDistChk_Config 2000 Answer 4
TTL_Config 5 SbrEvent 4
VplmnORCst_Config 1 AsyncEvent 1

DSR 8.4 Cloud Benchmarking Guide 34

The following is the MPS, CPU, CM description for a 10K MPS DA-MP VM profile.
Table 23: 10K MPS DA-MP VM Profile
Counter Measure (CM) MPS (K) CPU Operating Mode
All Stateful 6.9 49 Detection only
All Stateless 7.2 45.53 Detection only
All stateless excluding AVP Instance and Spec AVP Scr 9.6 44.6 Detection only
ApplicationId_And_Command_Code_Consistency_Check 10 35.84 Detection only
ApplicationId_Whitelist_Screening 10 36 Detection only
AVP_Multiple_Instance_Check 10 38.4 Detection only
Destination_Realm_Origin_Realm_Match_Check 10 35.94 Detection only
Origin_Host_And_Origin_Realm_Consistency_Check 10 35.86 Detection only
Realm_And_IMSI_Consistency_Check 10 36.3 Detection only
Origin_Realm_And_Destination_Realm_Whitelist_Screening 10 35.92 Detection only
VisitedPLMNId_And_Origin_Realm_Consistency_Check 10 36 Detection only
Specific_AVP_Screening 10 46.26 Detection only
Subscriber_Identity_Validation 10 37.26 Detection only
Previous_Location_Check 10 51.04 Detection only
Message_Rate_Monitoring 10 36.64 Detection only
Source_Host_Validation_Hss 10 50.81 Detection only
Source_Host_Validation_Mme 8.2 47.67 Detection only
Time_Distance_Check 9 44.44 Detection only

Number of concurrent sessions a SBR-u can handle is memory dependent. With Benchmarking, it is
observed that we can have up to 5Million records in SBR-u before it reaches 70% of system RAM with
record size of ~2K Bytes.

Service Capability Exposure Function (SCEF) Benchmarking

SCEF/MTC-IWF functionality at DSR adds two functional components.
• The Core SCEF/MTC-IWF component which provides interfaces towards the HSS, MME/SGSN,
PCRF. This functionality is part of DA-MP where SCEF/MTC-IWF application is implemented.
• The API gateway (OCSG) component which comprises of API admin, API App Server, API gateway
DB which provides exposure to SCS/AS on the HTTP interface.

Topology

DSR 8.4 Cloud Benchmarking Guide 35

 Figure 10: SCEF Testing Topology
There are three factors that determine the sizing of SCEF. The Performance of DA-MP with SCEF, API
gateway (OCSG) and the SBR universal (u). SBR- u is used by SCEF to store IoT Device context
information related to MTC procedures. In general, SBR-u capacity computation is based on number of
devices and number of MTC procedures planned to be deployed by MNO.
Number of U-SBR Sessions = Number of IoT Devices x Number of MTC procedures planned for
deployment.
Each of supported MTC procedures adds a separate record into U-SBR database.
Number of MPs and U-SBR server group’s required must be decided based on MNO’s traffic pattern and
IoT Device capacity. The capacity numbers are mentioned in section Summary of Benchmark Data
Recommendations.
The benchmarking involved running a call mix on a 5K MPS profile, of equal proportions for Non IP Data
Delivery, Monitoring and Device Trigger on a 4 DA-MP, 4 OCSG (API gateway) system to make sure all
memory, CPU, measurements and metrics stay in the acceptable range for MP's.
Key metrics for managing the SCEF DA-MP servers are:
Table 24: SCEF DA-MP Server Metrics
Recommended Usage
Measure-
ment ID Name Group Scope Description Condition Actions
19300 DxNiddMsgRateMetric SCEF System The number of messages If the Contact
(blade) processed every second by the numbers Oracle
NIDD feature of SCEF application exceed the
benchmark
19400 DxDevTriggMsgRateMetric SCEF System The number of messages
ed capacity
(blade) processed every second by the
Device Trigger feature of SCEF
application
19350 DxMonMsgRateMetric SCEF System The number of messages
(blade) processed every second by the
Monitoring feature of SCEF
application

Suggested Resolution
The growth mechanism for DA MPs, OCSG App Server is via horizontal scaling. That is by adding
additional DA MPs. The current maximum number of the DA MPs, OCSG App server per DSR signaling
NE is 32. If SCEF context information or MPS capacity is required then additional Server Groups may be
added to an existing SBR(u) using the SBR reconfiguration feature. There can be up to 64 Server
Groups in the SBR(u).

Equipment Identity Register (EIR)

The EIR application is enabled on the DA-MP with interface towards the MME/GGSN and STP MP
provides the SS7 Interface towards the MSC. An internal interface is provided as a query interface to the
UDR-NO.
The UDR NO provides the functionality of the Equipment Identity Register (EIR) database to the DSR.
The database stores white, gray, and black lists of IMEI numbers. It can support up to 100 Million
subscribers, 10 Million IMEI range/TAC with 1 IMEI supporting up to 10 IMSIs.
Topology

DSR 8.4 Cloud Benchmarking Guide 36

 Figure 11: EIR Testing Topology
Benchmarking was done for 2K EIR traffic using vSTP MP profile and DAMP profile, for vSTP EIR and
Diameter EIR respectively with 10 million subscribers configured and a max 500 TPS subscriber
provisioning traffic at UDR-NO. Each subscriber is configured up to 10 IMSIs. It was observed that the
Average response time of ~7ms UDR-NO Query processing time.

NOAM
Overview
Specific benchmark data for the DSR NOAM is not provided in this release as the DSR Cloud deployable
footprint is modest and system testing of the DSR indicates that NOAM growth in not currently needed.

Indicative Alarms/Events
The DSR Network OAM is potentially a RAM intensive function. The Network OAM is designed not to
exceed the available memory; however RAM is the most likely resource constraint.

Measurements
Measuring Network OAM Utilization
In this section, only the key recommended metrics for managing the performance of the Network OAM
are discussed. There are many more measurements available, and these can be found in [[2]].
The key metric for managing the Network OAM Servers are:
Table 25: Network OAM Metrics

Measure- Recommended Usage

ment ID Name Group Scope Description Condition Actions
31056 RAM_UtilPct_Average System System The average committed If the average Contact
(server) RAM usage as a Ram utilization Oracle
percentage of the total exceeds 80%
physical RAM utilization

Suggested Resolution
The NOAM can be vertically scaled; however this action is not anticipated to be necessary with the DSR
cloud deployable footprint. Please contact Oracle support for additional guidance as needed.

DSR 8.4 Cloud Benchmarking Guide 37

 SOAM
Overview
Specific benchmark data for the DSR SOAM is not provided in this release as the DSR Cloud deployable
footprint is modest and system testing of the DSR indicates that SOAM growth in not currently needed.

Indicative Alarms/Events
A key metric for managing the System OAM VM is:
Table 26: System OAM Metrics

Measure- Recommended Usage

ment ID Name Group Scope Description Condition Actions
31056 RAM_UtilPct_Average System VM The average committed If the average Contact
RAM usage as a Ram utilization Oracle
percentage of the total exceeds 80%
physical RAM utilization

Suggested Resolution
The DSR SOAM can be vertically scaled, the criteria for vertically scaling to a SOAM large profile (refer
DSR VM Configuration section) is set as > 2000 connections. Horizontal scaling of the DSR SOAM is not
supported or indicated in this release. Please contact Oracle support for additional guidance as needed.

IPFE
Overview
The IPFE was exercised in both VMware and KVM environments. Table 27 shows the measurement
capacity of the IPFE. Note that there are three main factors that determine the throughput limits:
• The number of TSAs (one or more) on the IPFE
• Whether there are more than 2,000 connections
• Whether the average message size is less than the MTU size.
Under most conditions the throughput of the IPFE is 2 Gbit/sec. However under the worst case of all
three of the above conditions the throughput of the IPFE drops to 1.6 Gb/sec.
When monitoring IPFE capacity both the guest and host CPU utilization should be monitored. Much of
the IPFE work is done at the host kernel level so the CPU utilization numbers returned by the IPFE
application level don’t fully reflect all of the IPFE overhead on the system.
Table 27: IPFE Throughput
Two or more TSAs on IPFE Pair
Single TSA on IPFE Pair (Total on both TSAs)
Avg Msg Size < Avg Msg Size >= Avg Msg Size < Avg Msg Size >=
1 MTU 1 MTU 1 MTU 1 MTU
2,000 Connections 2 Gbit/sec 2 Gbit/sec 2 Gbit/sec 2 Gbit/sec
or less
More than 2,000 2 Gbit/sec 1.6 Mbits/sec 2 Gbit/sec 2 Gbit/sec
Connections

DSR 8.4 Cloud Benchmarking Guide 38

 Topology

MME Simulator IPFE

IPFE DA-MPs
DA-MPs HSS Simulator

Figure 12. IPFE on Ingress Side Only

Figure 12 shows the typical IPFE configuration, with IPFEs used for “ingress” side traffic between the
Diameter clients. This configuration is typical since there are typically many more Diameter clients such
as MMEs than there are Diameter servers such as HSSs. The bandwidth numbers given in Table 27 are
for the traffic flowing between the clients (MME in the figure above) and the DSR DA-MPs.

MME Simulator IPFE

IPFE DA-MPs
DA-MPs IPFE
IPFE HSS Simulator

Figure 13: IPFE on both Ingress and Egress Sides

Figure 13 shows another possible IPFE configuration where the IPFE uses connection “initiator”
functionality to set up the connections between the DA-MPs and the Diameter servers (the HSS simulator
in this figure). The IPFE bandwidth requirements need to be calculated separately for the egress side. If
all of the DSR traffic goes through an IPFE on both ingress and egress sides, then the required IPFE
bandwidth is:
(2 * MPS) * (average Diameter message size including IP overhead))

Indicative Alarms/Events
In this section, only the key recommended metrics for managing the performance of the IPFE are
discussed. There are many more measurements available on the IPFE, and these can be found in [[2]].
Measurements
The key metrics for managing the IPFE VMs are:
Table 28: IPFE Metrics
Recommended Usage
Measure-
ment ID Name Group Scope Description Condition Actions
5203 RxIpfeBytes IPFE Server Bytes received by If the number of (bytes * 8 If the traffic is expected
Performance Group the IPFE bits/byte)/(time interval in s) is > to grow then, consider
benchmarked capacity (Gbps) adding an additional
IPFE pair
31052 CPU_UtilPct_Average System System The average CPU When running in normal Contact Oracle
(IPFE) usage from 0 to operation with a mate in normal
100% (100% operation, and this
indicates that all measurements exceeds 30%
cores are occupancy,
completely busy) OR
Exceeds 60% occupancy when
running without an active mate.
31056 RAM_UtilPct_Average System System The average If the average Ram utilization Contact Oracle
(IPFE) committed RAM exceeds 80% utilization
usage as a
percentage of the
total physical RAM

Suggested Resolution
Horizontal scaling by adding up to two pairs in total IPFEs per DSR Signaling NE as indicated.

DSR 8.4 Cloud Benchmarking Guide 39

 IDIH
Overview
The IDIH (IDIH application, IDIH mediation, IDIH Database VMs) are considered a best effort trouble
shooting tool for the DSR. Benchmarking data is not currently provided for the IDIH VMs.

Suggested Resolution
Contact Oracle support.

Appendix A. DSR VM Configurations

The information shown below is a summary of the VM configurations used for the benchmarking data,
and the affinity rules for deploying those VMs. Using VM sizes different from these tested values may
give unexpected results since the application profiles are tuned to this number of vCPUs and memory
sizes. In particular the disk size indicated in Table 26 as storage for each VM is an absolute minimum
required to load the DSR images.
Table 29: VM Configurations and Affinity Rules

RAM Disk Affinity/Placement

VM Name vCPU (GB) (GB) Max Config Redundancy Models Rules (Per Site) Notes
DSR 4 6 70 1 Pair Active/Standby 2 VMs per DSR
NOAM network in any site.
(Regular) VMs to be deployed
on separate servers
DSR 8 14 70 1 Pair Active/Standby 2 VMs per DSR It is recommended to use large NOAM
NOAM network in any site. profile if the deployment is more than 32
(Large) VMs to be deployed C level servers.
on separate servers If SOAM profile is large, NOAM profile
must be large. In large profiles, the
scheduled two measurement exports will
run in parallel.
DSR 4 6 70 1 Pair per DSR Active/Standby 2 VMs per site. VMs Redundancy model
SOAM NF or to be deployed on Active/Standby/Spare model is used for
(Regular) Active/Standby/Spare separate servers. PCA mated-pair deployments. For all
other deployments Active/Standby model
is used.
DSR 8 14 70 1 Pair per DSR Active/Standby 2 VMs per site. VMs Redundancy model
SOAM NF or to be deployed on Active/Standby/Spare model is used for
(Large) Active/Standby/Spare separate servers. PCA mated-pair deployments. For all
other deployments Active/Standby model
is used.
If the diameter connections are higher
than 4K, it is recommended to use SOAM
large profile.
If NOAM profile is large, SOAM profile
must be large. In large profiles, the
scheduled two measurement exports will
run in parallel.

DSR 8.4 Cloud Benchmarking Guide 40

 RAM Disk Affinity/Placement
VM Name vCPU (GB) (GB) Max Config Redundancy Models Rules (Per Site) Notes
DA MP 12 16 70 32 per DSR NF Active Cluster (N+0) Should be spread The limit of 32 is the combined total of
over as many servers DA-MPs, DA-MPs with IWF, DA-MP with
as possible to EIR and DA-MPs with SCEF. The vSTP
minimize capacity MPs and SS7 MPs do not count against
loss on server loss this 32. Cannot max out all types in one
DSR (for instance 32 DA-MPs AND 32
vSTPs). Any solution using more than
500 ART(Application Route
Tables)/ARR(Application Route
Rules)+PRR(Peer Route Rules) beyond
20k please use the below profile(DAMP
w/ IWF) which is with 24 GB RAM.
DA MP 12 24 70 32 per DSR NF Active Cluster (N+0) Should be spread The 24 GB RAM requirement is a hard
w/IWF over as many servers minimum if the DA-MP VM is used with
as possible to the IWF application.
minimize capacity
loss on server loss
vSTP MP 8 8 70 32 per DSR NF Active Cluster (N+0) Should be spread The vSTP MPs do not count against the
over as many servers 32 DA-MP limit in a single OCDSR node,
as possible to so a DSR can have up to 32 vSTP MPs.
minimize capacity Cannot max out all types in one DSR (for
loss on server loss instance 32 DA-MPs and 32 vSTPs).
IPFE 6 16 70 2 pairs per DSR Active/Standby Each VM in a pair Deployed in pairs. Max 2 pairs (4 VMs).
NF must be deployed on
separate server
SS7 MP 12 24 70 8 per DSR NF Active Cluster (N+0) Should be spread
over as many servers
as possible to
minimize capacity
loss on server loss
SBR(s) 12 25 70 8 Server Groups Active/Standby/Spare Active/Standby VMs Can be either Active/Standby/Spare or
per SBR(s) to be deployed on Active/Standby depending on customer
separate servers, geo-redundancy requirements.
Spare is typically at
another geographic
location for geo-
redundancy
SBR(b) 12 32 70 8 Server Groups Active/Standby/Spare Active/Standby VMs Can be either Active/Standby/Spare or
per SBR(b) to be deployed on Active/Standby depending on customer
separate servers, geo-redundancy requirements.
Spare is typically at
another geographic
location for geo-
redundancy
SBR(u) 12 24 70 64 Server Active/Standby/Spare Active/Standby VMs Can be either Active/Standby/Spare or
Groups per to be deployed on Active/Standby depending on customer
SBR(b) separate servers, geo-redundancy requirements.
Spare is typically at
another geographic
location for geo-
redundancy
DP SOAM 4 12 125 1Pair per DSR Active/Standby 2 VMs per site. VMs
(Regular) NF to be deployed on
separate servers
DP 6 10 125 10 per DSR NF Active Cluster (N+0) Should be spread To be evenly distributed across servers
(Regular) over as many servers to minimize capacity loss
as possible to
minimize capacity
loss on server loss

DSR 8.4 Cloud Benchmarking Guide 41

 RAM Disk Affinity/Placement
VM Name vCPU (GB) (GB) Max Config Redundancy Models Rules (Per Site) Notes
SDS 4 32 300 1 Pair per Active/Standby Anti-affinity between Active/Standby. An optional "Disaster
NOAM Network the Active/Standby Recovery" SDS is supported that would
(Regular) VMs typically be located at a different data
center to provide geo-redundancy.
Query 4 32 300 1 per SDS N/A since non- Non, non-redundant Optional 1 per site. Can have one for the
Server NOAM redundant primary SDS-NOAM and one for the
(Regular) Disaster Recovery SDS-NOAM
DP SOAM 4 64 400 1Pair per DSR Active/Standby 2 VMs per site. VMs Supports
(Large) NF to be deployed on
separate servers
DP 24 64 400 10 per DSR NF Active Cluster (N+0) Should be spread To be evenly distributed across servers
(Large) over as many servers to minimize capacity loss
as possible to
minimize capacity
loss on server loss
SDS 4 128 840 1 Pair per Active/Standby Anti-affinity between Active/Standby. An optional "Disaster
NOAM Network the Active/Standby Recovery" SDS is supported that would
(Large) VMs typically be located at a different data
center to provide geo-redundancy.
Query 4 128 840 1 per SDS N/A since non- Non, non-redundant Optional 1 per site. Can have one for the
Server NOAM redundant primary SDS-NOAM and one for the
(Large) Disaster Recovery SDS-NOAM
UDR NO 14 64 400 n Active/Standby/Spare Active/Standby/Spare Redundancy model
(Active,Standby) VMs to be deployed Active/Standby/Spare model is used.
on separate servers, Active/Standby on Site 1 and Spare on
Spare is typically at Site 2. UDR is scaled by adding
another geographic UDRNOs. The Standby UDR NO also
location for geo- receives query traffic from STP-MP and
redundancy DA-MP.
API 12 16 70 32 per DSR NF Active Cluster (N+0) Should be spread Optional component.
over as many servers
Gateway as possible to For SCEF deployments only.
App minimize capacity
Server loss on server loss.

API 4 6 70 1 per DSR NF Active None, non-redundant Optional component.

Gateway For SCEF deployments only.
Admin
Server
API 4 6 70 1 pair per DSR Active/Standby Active/Standby VMs Optional component.
NF to be deployed on
Gateway separate servers. For SCEF deployments only.
DB server
VNFM 4 6 70 NA N/A since non- None, non-redundant VNF Manager deployment
redundant
SPF+NRF 2 4 30 NA N/A since non- None, non-redundant Optional component.
redundant 5G apps deployment only.
iDIH 4 8 64 1 per Site N/A since non- None, non-redundant Optional component for Diameter traffic
Application redundant monitoring
iDIH 4 8 64 1 per Site N/A since non- None, non-redundant Optional component for Diameter traffic
Mediation redundant monitoring
iDIH 4 8 200 1 per Site N/A since non- None, non-redundant Optional component for Diameter traffic
Database redundant monitoring

DSR 8.4 Cloud Benchmarking Guide 42

 Appendix B. DSR VM Disk Requirements
This section provides guidance on the disk requirements for the OCDSR VMs. Characterizing disk
requirements can be tricky since there are many variables that can affect disk usage, such as how many
reports are being run on the OAM systems, or how often backups are run. Peak disk utilization can also
very different from average disk utilization, for instance during backups or restore operations. While
these guidelines are provided for the disk usage of the different VM types, customers should verify their
disk usage under their own conditions since they are more driven by how the customer uses their system
than by easier to calculate factors such as CPU utilization per MPS.
The OCDSR has been designed as a low disk-utilization application, with all critical call processing
applications performed in memory. There is also no swap disk utilization in any of the VMs. As a
background for all of these numbers, the OCDSR has been run for years on “bare metal” deployments
with a single pair of industry-standard 10k RPM, 2½ inch disk drives in Raid 1. So even maximum sized
OCDSR configurations run successfully on the approximately 120 IOPs provided by those disks. When
run on higher-performance disk subsystems such as SSDs, high disk utilization tasks such as
background report generation just complete faster. The notes for each VM type give some of the factors
that can drive different disk utilization levels. For instance, the primary traffic handling VMs, the IPFEs
and the different types of MPs, have a fairly constant disk utilization independent of the traffic level. This
is because the primary disk utilization is for saving statistics, then forwarding them to the SOAM.
Table 30: VM Disk Utilization Characteristics
Routine Disk Peak Disk
Disk Utilization Utilization
VM Name (GB) (IOPs)1 (IOPs)2 Disk Usage Modes Notes
DSR NOAM 70 100 800 Periodic (30 second) Background disk utilization is mostly statistics
small writes to collect collection from managed DSRs. Peak disk
DSR SOAM 70 100 800
statistics. utilization driven by customer report generation
Large block reads to run and maintenance activities such backups.
reports and backups.
DA MP 70 50 500 Writes statistics to disk Disk utilization is independent to traffic levels.
at 30 second intervals, Is affected by the size of the DSR configuration
DA MP 70 50 500
w/IWF reads them at 5 minute (number of connections for instance) and the
intervals to send to utilization of features that create measurements
SOAM such as ETGs and TTPs.
vSTP MP 70 50 500 Disk utilization is independent to traffic levels.
Is affected by the size of the vSTP configuration
such as the number of local and remote peers.
SS7 MP 70 50 500 Disk utilization is independent to traffic levels.
Is affected by the size of the DSR configuration.
IPFE 70 20 100 IPFE has relatively few configuration items and
statistics, and very low disk utilization.
SBR(s) 70 50 800 Disk writes mostly short Peak disk utilization driven by recovery
bursts for statistics activities between active/standby servers.
SBR(b) 70 50 800
storage
SBR(u) 70 50 800
SDS NOAM 300 100 800 Synchronizes changes SDS can maintain multiple copies of large
to in-memory database subscriber database. Peak disk utilization is
to disk. Mostly write mostly driven by creating new backups.
application
DP SOAM 125 50 800
DP 125 80 500

DSR 8.4 Cloud Benchmarking Guide 43

 Routine Disk Peak Disk
Disk Utilization Utilization
VM Name (GB) (IOPs)1 (IOPs)2 Disk Usage Modes Notes
Query 300 100 800 Synchronizes changes The query server is not a real-time system. The
Server to in-memory database amount of disk reads is driven entirely by
to disk. Reads are manual customer database queries.
driven by customer
queries
1 The“routine” disk utilization is the minimum engineered IOPs for the proper functioning of the VM.
Average disk utilization is typically lower.
2 The “Peak” disk utilization is number of IOPs the VM is capable of using given sufficient resources.

Appendix C. VM Networking Requirements

This section gives information on the networking characteristics of the different VMs. The traffic is broken
down into signaling traffic handled on the XMI network, and OAM traffic carried on the IMI and XMI
networks.
The Diameter Traffic requirements on the XSI networks can be calculated from the MPS. Treating the
OCDSR as black box, this network traffic is simply the average Diameter message size (for requests and
answers) times the MPS rate for the OCDSR node. The complication is that some Diameter traffic is
likely to go through both an ingress DA-MP and an egress DA-MP. The most conservative consumption
is that any ingress message is equally likely to go out any of the DA-MPs. Thus if a DSR has X DA-MPs,
and Y total MPS, the average Diameter signaling traffic through a DA-MP is:
((Average Diameter message size including IP overhead) * Y) * (1+ ((2X-1)/X))
As an example, if the average Diameter message size is 2,000 bytes including overhead, the overall DSR
MPS is 10,000 MPS, and the number of DA-MPs is three, the calculation would be:
(2,000 bytes * 8 bits/byte *10,000 MPS) * (1+ (2*3 DA-MPs) -1)/(3 DA-MPs)) = (160,000 kb/s) * (2.66) = 426,666 kb/s per DA-MP
For the MP types other than the DA-MPs simply substitute the average size of signaling types, for
instance SS7 messages for the vSTP MP. Since typically SS7 messages are much smaller than
Diameter messages (for instance ~200 bytes for SMS), the vSTP MP bandwidth is much smaller than the
DA-MP bandwidth.
The OAM traffic on the VMs can be much more variable since it’s dependent to customer-specific usage
patterns such as the number of reports requested and the number of periodic activities such as backups
and restores. The notes for each VM type give some background on the network impacts of these
customer-driven activities.
Table 31: VM Networking Utilization Characteristics
Management Traffic
Networks Networks Networks
VM Name Supported (Gb/s) (Gb/s) Notes
DSR XMI 2 N/A Activities such as backups can generate higher
NOAM IMI network utilization, but runs at the rate of the
DSR 1 bandwidth available since they are not real-time
SOAM activities.

DA MP XMI 0.2 MPS See explanation above for how to calculate the
DA MP IMI Dependent signaling network traffic.
w/IWF XSI
vSTP MP
SS7 MP

DSR 8.4 Cloud Benchmarking Guide 44

 Management Traffic
Networks Networks Networks
VM Name Supported (Gb/s) (Gb/s) Notes
IPFE XMI 0.2 MPS The peak networking capacity supported by the IPFE is
IMI Dependent 3.2 Gb/s. Typically the IPFE is deployed only on the
XSI ingress (towards clients such as MMEs) side of the
DA-MP, so the total traffic through the IPFE is ½ the
total bandwidth of the DA-MPs.
SBR(s) XMI 1.0 N/A The given OAM bandwidth is for routine operations.
SBR(b) IMI Some recovery operations such as synchronizing the
database between the active and standby servers after
SBR(u) a prolonged disconnection can consume an order of
magnitude or more of network bandwidth. The
required amount of bandwidth for these recovery
operations is very dependent on customer-factors such
as number of subscribers, the MPS rate, and the
amount of networking downtime.
SDS XMI 1.0 N/A The maximum bandwidth required by the SDS NOAM
NOAM IMI is determined primarily by the provisioning rate from
external customer systems along with the size of the
customer records.
DP SOAM XMI 1.0 N/A All of the subscriber data provisioned at the SDS
IMI NOAM is passed down to each DP SOAM, which then
distributes the data to any attached DPs.
DP XMI 1.0 N/A The DP receives writes of new subscriber records from
IMI the SOAM, and database queries from the DA-MPs.
Query XMI 1.0 N/A The Query Server is synchronized to the changes in
Server IMI the SDS NOAM. In addition there is some network
traffic due to customer search requests, but this traffic
is small compared to the synchronization traffic.
UDR NO XMI 1.0 N/A UDR NO receives internal query from STP MP and DA
IMI MP
XSI

Table 32 shows some guidelines for mapping the logical OCDSR networks (XMI, IMI, etc.) to interfaces.
There is nothing fixed about these assignments in the application, so they can be assigned as desired if
the customer has other requirements driving interface assignment.
Table 32: Typical OCDSR Network to Device Assignments
VM OAM Local Signaling Signaling Signaling Signaling Signaling Replication DIH
Name (XMI) (IMI) A (XSI1) B (XSI2) C (XSI3) (…) D (XSI6) (SBR Rep) Internal
DSR eth0 eth1
NOAM
DSR eth0 eth1
SOAM
DA-MP eth0 eth1 eth2 eth3 eth4 eth17 eth18
IPFE eth0 eth1 eth2 eth3 eth4 eth17
SS7 MP eth0 eth1 eth2 eth3 eth4 eth17 eth18
SBRB eth0 eth1 eth2
SBRS eth0 eth1 eth2
SBRU eth0 eth1 eth2

DSR 8.4 Cloud Benchmarking Guide 45

 VM OAM Local Signaling Signaling Signaling Signaling Signaling Replication DIH
Name (XMI) (IMI) A (XSI1) B (XSI2) C (XSI3) (…) D (XSI6) (SBR Rep) Internal
vSTP eth0 eth1 eth2 eth3 eth4 eth17
UDRNO eth0 eth1 eth2 eth3 eth4 eth17
iDIH App xmi int
iDIH Med xmi imi int
iDIH DB xmi Int
SDS eth0 eth1
NOAM
SDS eth0 eth1
NOAM
DP eth0 eth1
Query eth0 eth1
Server
API eth0 eth1
Gateway
Admin
(OCSG)
API eth0 eth1 eth2 eth3 eth4 eth17
Gateway
App
Server
(OCSG)
API eth0 eth1
Gateway
DB
(OCSG)

Appendix D. Summary of Benchmark Data Recommendations

The information shown below is a summary of the benchmark data described throughout the document.
This data is intended to provide guidance, and is based solely on the observed results from the test
setups described in this document. Recommendations may need to be adapted to the conditions in a
given operator’s cloud, such as differences in traffic patterns, feature utilization patterns, and
infrastructure differences.
Table 33: Benchmark Data Summary
Benchmark Run VMware/ESXi Openstack/KVM
Application Software DSR 8.3 (running Oracle Linux) DSR 8.3 (running Oracle Linux)
Host VM VMware 6.0 OpenStack Mitaka, Newton/KVM
HW Oracle Server X5-2 Oracle Server X5-2, HP Gen9v1
VM Profiles/Flavors DSR VM Configurations DSR VM Configurations

Table 34: Recommended Maximum Engineering Targets

Recommended Maximum Engineering Targets
VM Name VM Purpose Unit Quantity
Network Operation,
DSR NOAM Administration, Maintenance VM 1+1
(and Provisioning)
Site (node/Network Element)
DSR SOAM Operation, Administration, VM 1+1
Maintenance (and Provisioning)

DSR 8.4 Cloud Benchmarking Guide 46

 Recommended Maximum Engineering Targets
VM Name VM Purpose Unit Quantity
Diameter Agent Message
DA MP (Relay) MPS 18,000
Processor
DA MP (Relay)
configuration
set to
DOC/CL1/CL2 Diameter Agent Message
discards set to MPS 36,000
Processor
0 and multi
queuing
enabled on all
hosts
DA MP Diameter Agent Message
MPS 16,000
(Database) Processor
DA MP Diameter Agent Message
MPS 13,000
(Statefull) Processor
Diameter Agent Message
DA MP w/ IWF MPS 12,000
Processor with IWF
Diameter Agent Message
DA MP w/ EIR MPS 2000
Processor for EIR application
DA MP w/ Diameter Agent Message
MPS 5000
SCEF Processor for SCEF application
Virtual STP for M3UA and M2PA
vSTP MPS 10,000
message Processing
Virtual STP Message processor
vSTP w/ EIR MPS 2000
with EIR application
Gb/s per IPFE pair 2.01
Connections per IPFE Pair per
2,000
IPFE IP Front End TSA
Target Set Address (TSA) per
32
IPFE pair
SS7 Message Processor for
SS7 MP MAP Diameter interworking MPS 12,000
function
SBR(s) (Single Subscriber Binding Repository Diameter sessions 16,000,000
Server Group) (session) for Policy DRA MPS 50,000
SBR(s) (Max 8 Subscriber Binding Repository Diameter sessions 128,000,000
Server Groups) (session) for Policy DRA MPS 400,000
SBR(b) (Single Subscriber Binding Repository Subscriber Bindings 16,000,000
Server Group) (binding) for Policy DRA MPS 50,000
SBR(b) (Max Subscriber Bindings 128,000,000
Network-wide Subscriber Binding Repository
with 8 Server (binding) for Policy DRA MPS 400,000
Groups)
Context Information 5,000,000

DSR 8.4 Cloud Benchmarking Guide 47

 Recommended Maximum Engineering Targets
VM Name VM Purpose Unit Quantity
SBR(u) (Single Subscriber Binding Repository
MPS 16,000
Server Group (universal) for SCEF
SBR(u) (Max Context Information 320,000,000
Subscriber Binding Repository
64 Server
(universal) for SCEF MPS 1,024,000
Groups)
Database Processor Site (node)
Operation, Administration,
DP SOAM Maintenance for address VM 1+1
resolution and subscriber
location functions
Database Processor for address
MPS requiring DP lookups
DP resolution and subscriber 80,000
(usually 50% of FABR traffic)
location functions
Routing Entities (typically 2x
300,000,000
subscriber count)
Subscriber Database Processor
SDS for address resolution and NAI User Routing Entries 500,000
subscriber location functions Provisioning TPS 200
XML Interface reads/second 200
Allows customers to query
Query Server FABR subscriber data via a N/A N/A
MySQL interface
User Data Repository for vEIR MPS 12,500
UDR NO
message Processing Max Number of Subscribers 100,000,000
1 See also Table 27 for factors that can lower this number.

DSR 8.4 Cloud Benchmarking Guide 48

Appendix E. Detailed Infrastructure Settings
Table 35: Detailed Infrastructure Settings

Attribute KVM/Oracle X5-2 HP Gen9v1

Model Oracle Server X5-2 ProLiant DL380c Gen9v1
Processor Intel® Xeon® CPU E5-2699 v3 @ Intel(R) Xeon® CPU E5-2680 v3 @
Type 2.30GHz 2.50GHz
vCPUs 72 [2 CPU Sockets [18 x 2 Cores, each 48 [2 CPU Sockets [12 x 2 Cores, each
with Hyper threading Active] with Hyper threading Active]
RAM 128 G [DDR4-2133] 256 GB (DDR4-2133)
CPU Cache 45 MB (Intel® Smart Cache) 30MB (Intel® Smart Cache)
Memory
Number and 4 [Intel Corporation Ethernet Controller 2 x HP 10Gb 2-port 560M Adapter
Type of NICs 10-Gigabit X540-AT2]
BIOS Power Power Supply Maximum: Maximum power HP Static High Performance Mode
Settings the available PSUs can draw
Allocated Power: Power allocated for
installed and hot pluggable components
Peak Permitted: Maximum power the
system is permitted to consume (set to
Allocated Power)
HDD 2.3 TB of solid state drive (SSD) storage HP Smart Array P244br Controller
Disk Drive Interface 6Gb/s SAS (Serial
Attached SCSI)
2 x HP 900GB 12G SAS 10K rpm SFF
(2.5-inch) SC Enterprise

DSR 8.4 Cloud Benchmarking Guide 49

Appendix F. Small DSR VM Configuration
Many customers don’t need the capacity provided by even a single pair of standard size DSR DA-MP
VMs. The fixed configuration in this section is for customers that only need:
• Relay or RBAR applications
• 6k MPS or less (assuming an infrastructure equal to or better than the X5-2 processors used for the
benchmark tests).
• 5K or less RBAR entries
• No IPFE
This configuration should be run with the 6k DA-MP profile.
Table 36: Minimal System VM Configurations and Affinity Rules

RAM Disk Max Redundancy Affinity/Placement

VM Name vCPU (GB) (GB) Config Models Rules (Per Site) Notes
DSR NOAM 4 6 70 1 Pair Active/Standby 2 VMs per DSR Two NOAMs are
network in any site. always required to
DSR SOAM 4 6 70 1 Pair Active/Standby
VMs to be support upgrades, but
per DSR
deployed on they can be on the
NF
separate servers if same server if only
possible one server is available.
DA MP 4 8 70 1 or 2 1 or 1+1 The 1+1 There’s little value in
per DSR configuration have redundant DA-
NF should have the MPs (1+1) if they are
DA-MPs on on the same server
different servers since the server is the
thing most likely to fail.

DSR 8.4 Cloud Benchmarking Guide 50

Appendix G. DSR DA-MP Profiles
DA-MP profiles are used to set various parameters associated with performance of the DA-MP such as:
• Internal resource allocations such as the number of various types of threads created by that VM.
• The absolute limit of traffic that a DA-MP can handle set to 2.5 times the stated capacity of the VM.
• Fixed alarms for various traffic conditions such as message rates, average message sizes and hold
times.
• User-configurable values for message discard percentages at the different DA-MP congestion levels.
The throughput of a DA-MP is determined by the CPU and infrastructure, not by the Profile assigned to it.
Assigning a higher profile does not increase the throughput of a DA-MP, and may hurt it. For perspective,
if you apply all of the factors below to the X5-2 system used in the benchmarks running 100% relay, you
get the recommended capacity of 18k MPS per DA-MP. Thus all of the benchmarks in this document
were run with the 18k profile.
In the cloud environment the correct profile should be selected based on the expected DA-MP
throughput. This throughput is based on three major factors:
• The performance of the customer’s infrastructure relative to the infrastructure used for the
benchmarking described in this document.
• The mix of application types (relay, database, session) being run on the DA-MP.
• The VM configuration such as the number of vCPUs.
These factors are useful for estimating the VM performance in advance, but the customer should also
monitor the actual performance of the different VM types using the KPIs, alarms, and measurements
described for each VM type in the previous sections.

Infrastructure Performance Differences

The benchmarks in this document were done on Oracle X5-2 servers configured with Intel Xeon X5-
2699v3 processors. When trying to estimate the performance of a different type of server with a different
processor type, the main factor is the single threaded performance of the chip, such as SPECint® 2006
from Spec.org. The overall performance of the server/chip is less important since the VM configurations
assign the same number of vCPUs/threads to the VM independent of the number of threads supported by
processor (for example, you can get more VMs on a server with more vCPUs, but the performance for a
given number of vCPUs is set by the single-thread performance). Historically, within a couple of Intel
processor generations, the performance of the OCDSR VMs varies relatively linearly with the difference in
single-thread performance, for example, a processor with a SPECint2006 rating of 66 has 10% better
OCDSR performance than a processor with a SPECint2006 rating of 60.

Traffic Mix
The different classes of applications tested (Relay, Database, Statefull) have significantly different MPS
results for a fixed infrastructure configuration (server type, VM size). The capacity for an OCDSR running
a mixture of these types can be calculated by using percentage of traffic of a given type. For example,
using the values from Table 33, consider an OCSR with the following traffic mix:
• Relay 40% (18k MPS)
• Database (30%) (16k MPS)
• Statefull (30%) (13k MPS)
The effective throughput for this traffic mix would be:
(40%*18k) + (30%*16k) + (30%*16k) =15.9k MPS

DSR 8.4 Cloud Benchmarking Guide 51

The caveat to this calculation is that the percentage for the lower performing traffic types (Database and
Statefull) should reflect their worst case values, for instance to handle traffic spikes of that traffic type
caused by failure conditions of other components in the network.
The performance of the vSTP application is calculated separately since it has dedicated MPs that run
only vSTP traffic.

VM Configuration
The tested VM configurations are given in Section DSR VM Configurations and Section DSR DA-MP
Profiles. It’s recommended that these VM configurations are used as tested for the most predictable
results. If for some reason the customer desires to change them for a given installation, here are some
guidelines that should be kept in mind:
• The installation and upgrade procedures for the VMs requires a minimum of 70GB of disk storage.
While assigning less storage than this may appear to work during installation, it will likely cause
failures during the upgrade procedures.
• The IWF function requires 24 GBs of memory to run. It does not come into service with any smaller
memory allocation.
• Adding vCPUs to the configurations may increase performance, but only up to a point since there
may not be enough threads to efficiently take advantage of the extra vCPUs.
• Reducing the vCPU counts should not be done for any VM except for the DA-MPs. The issue is that
configurations that appear to run fine under normal traffic conditions may not have sufficient capacity
under recovery conditions or under heavy loads (for instance running reports on a SOAM while it’s
doing a backup). The DA-MP vCPU number should not be lower than 4 vCPUs (see section Small
DSR VM Configuration for an example small DSR configuration.)

Selecting a DA-MP Traffic Profile

The following DA-MP traffic profiles for cloud deployments are supported in OCDSR 8.4:
• 6k MPS
• 8k MPS
• 12k MPS
• 14k MPS
• 16k MPS
• 18k MPS
• 21k MPS
• 24k MPS
• 27k MPS
• 30k MPS
It is possible to change the DA-MP profile assigned to a DA-MP VM, however, the DA-MP application has
to be restarted for it to take effect. Since the results of the traffic mix calculation and any adjustments due
to different processor performance are unlikely to land exactly on any of these numbers, the question is
whether to use the next larger or next smaller DA-MP profile. In general, selecting the next larger DA-MP
profile (for instance selecting the 16k MPS profile for the 15.9k MPS calculated in the Traffic Mix) provides
the best solution. However, here are some considerations:
• While the larger profiles assign more system resources (message buffers, etc.) the differences
between these resource allocations between any two consecutive profiles is relatively small

DSR 8.4 Cloud Benchmarking Guide 52

 (proportional to their MPS difference or about typically about 20%). Since these allocated system
resources are a relatively small portion of the over VM memory usage (which includes, for instance,
the resources used by the Guest OS and the base OCDSR application), assigning the higher rated
profile ensures sufficient resource allocation. The standard VM profiles provided in DSR VM
Configurations section have sufficient memory to support up to the 24k MPS profiles. Above that
level, additional memory may be required depending on the traffic mix.
• Over a small range of DA-MP profiles (plus or minus one at least), the DA-MP profile likely has no
impact on the actual throughput of the VM. The DA-MP overload controls are driven by CPU
percentage (that is, actual utilization), not by any of these calculations. The only hard limit on the
throughput of a given profile is 250% of the nominal value. For instance, the 10k MPS DA-MP profile
starts shedding traffic (using the same algorithms as DA-MP overload) at 25k MPS, even if the CPU
percentage has not hit the CPU utilization level required to trigger the CL1 congestion level.
• Selecting the DA-MP profile just under the calculated MPS capacity (for example, selecting the 14k
DA-MP profile for the 15.9k MPS calculated in the Traffic Mix) causes the various system alarms to
happen earlier. For instance, the minor traffic alarm triggers at 8400 MPS (60% of 14k) instead of at
9200 MPS (60% of 16k) if the next larger profile was selected. Thus selecting the smaller DA-MP
profile (14k, in this example) provides more conservative alarming, and the higher profile (16K MPS)
provides less conservative alarming. But, as noted in the previous item, it does not affect the actual
throughput of the VM since that is driven by CPU usage.
• Selecting a DA-MP profile that is a lot larger than the nominal capacity calculated in the previous
section does not increase the throughput of the VM, and causes the capacity alarms to trigger later
than expected. Selecting a DA-MP profile that is a lot smaller than the nominal capacity causes the
alarms to trigger sooner than expected, and may limit the capacity of the VM due to insufficient
system resources being allocated.

DSR 8.4 Cloud Benchmarking Guide 53