AMS/WMS

MICROCODE VERSION 0740/C

RELEASED 05/26/2006

Newly supported features and functions for Version 0740/C

1) Shadow Image support for AMS/WMS/NAS
Item Description
Contents Shadow Image function for AMS/WMS/NAS Modular is supported with NAS OS 04-01
2) Maintenance IP Auto-swap without reboot function support
Item Description
Contents In Auto-swap mode, when IP address of User port is changed, IP address of Maintenance port
automatically changed into following address without reboot.
- 10.xxx.xxx.xxx --- CTL0 : 192.168.0.16, CTL1 : 192.168.0.17
- other than above --- CTL0 : 10.0.0.16, CTL1 : 10.0.0.17
3) LAN setting with Constitute file
Item Description
Contents User port and Maintenance port LAN Setting function with a constitute file has been added.
With this function, you can store the LAN constitution of the system to a file, and apply the
LAN constitution into another subsystem without reboot.
4) Mozilla 1.5 Browser support
Item Description
Contents We support Mozilla 1.5 browser for following OS:
AIX 4.3 and AIX 5.1

Improvements for Version 0740/C

1) NAS Performance monitoring function support
Item Description
Contents Performance monitoring function is supported when NNC is attached to AMS/WMS
2) Statistical information about host command logged in trace information
Item Description
Contents 2) Statistical information about host command is logged in trace information
3) IP connection check before starting IP change
Item Description
Contents We support IP connection check before changing IP address without reboot.
If any connection on the LAN ports remain when changing its IP address, then the IP address
change sequence without reboot will stop and error message displays.
Furthermore, information about the connecting IP address on the LAN ports is output to
“netstat.inf” file in the directory where Storage Navigator Modular installed.
4) IP connection check before starting IP change
Item Description
Contents QLA2460 FC HBA support
5) ISCSI support platform enhancement
Item Description
Contents A) Microsoft ISCSI initiator 2.01 supported
B) Red Hat Linux with NIC is supported

HDS Confidential 1 of 7
The change of contents for Version 0740/C

1) Host I/O may fail while using a spare disk drive

Item Description
Contents When a spare HDD is used (refer to Conditions Occurrence (B) Spare disk drive is used), the
LU access may fail because controller responds to a host I/O with Check Condition status
(Sense Key = 0Bh, Sense Code = C006h). Interruption occurs due to microcode bug in logic to
check write data for spare disk drive after write processing to spare disk drive and write data for
spare drive may not write to spare disk drive. As result, cache memory of controller busy with
write data for spare disk drive, and the host I/O will not execute because of cache segment
allocated.
With all of the following conditions met, the phenomenon may occur.
A) Microcode version is 0730 or later
B) The spare disk drive is used:
- Copy-Back-Less mode then from beginning to end of correction copy operation to spare drives
- Copy-Back mode then from the beginning of the correction copy operation to the spare disk
drive, to the end of the copy back operation to the replaced disk drive
C) The host issues I/Os only to the RAID group using the spare disk drive
(The write processing to the spare disk drive works when host I/Os are not issued to the RAID
group or host I/Os are issued to the other RAID group)

2) Controller blocked when displaying LUN Manager Information on WEB

Item Description
Contents When displaying the LUN Manager Information on WEB, a controller failure may occur.
The problem may occur when all of the following conditions are met:
A) LUN Manager is enabled
B) LU Mapping Mode is enabled
C) Host Group Security is enabled
D) More than 54 WWN are registered

3) PSUE occurred with setting Host Group on SNM

Item Description
Contents When the Host Group Setting changed on SNM, the True Copy pair may become PSUE. Due to
a microcode bug in re-login sequence after WWN setting on Host Group Security, PSUE
occurred by taking unexpectedly long time to re-login.
The problem may occur when all of the followings occur at the same time.
A) LUN Manager is enable on the port which True Copy is using
B) Host Group Security is enable
C) Setting WWN Information with SNM

4) Clearing the port configuration not completed on the change of the interface board
Item Description
Contents When the type of the installed interface board is changed, the port option setting does not clear,
due to a microcode bug in the initialization process.
In changing interface board, the following information should be cleared.
A) Host Group
B) LU Mapping
C) Port Options
Setting of Host Group and LU Mapping are cleared normally, but Port Options are not cleared.

HDS Confidential 2 of 7
5) True Copy Path failure occurred with "command time out"
Item Description
Contents If sequential write to the True Copy P-VOL would be done, the path failure will occur. Due to a
microcode bug in drive writing process, timeout may occur by taking unexpected long time to
access the larger size stripe.
The problem may occur with the following condition on True Copy S-VOL:
A) RAID 6, Stripe size:512KB, Segment size 16KB, 5D+2P and more
B) RAID 6, Stripe size:512KB, Segment size 64KB, 18D+2P and more
C) RAID 6, Stripe size:256KB, Segment size 8KB, 5D+2P and more
D) RAID 6, Stripe size:256KB, Segment size 16KB, 13D+2P and more
E) RAID 6, Stripe size:64KB, Segment size 4KB, 13D+2P and more
F) RAID 5, Stripe size:512KB, Segment size 16KB, 8D+1P and more
G) RAID 5, Stripe size:256KB, Segment size 8KB, 8D+1P and more

6) True Copy Path failure occurred with "Path login failed”

Item Description
Contents In the True Copy configuration via FC switch, when one subsystem reboot, "Path login failed"
occurs and the path, do not recover automatically. Due to a microcode bug in path login/logout
process after one system reboot, one of the paths may fail to recover.

7) Controller failure with "Data transfer check error [CCPLUNERR]"

Item Description
Contents Controller failure may occur with the following message using Shadow Image. Web Message:
Data transfer check error [CCPLUNERR], due to a microcode bug in the "Write & Verify
command" processing for SVOL of Shadow Image.
With all the following conditions met, the problem may occur.
A. Shadow Image is used
B. The pair status of Shadow Image is "COPY" or "PAIR"
C. Subsystem receives "Write & Verify command” from a host for SVOL of Shadow Image.

8) Fibre channel Link Down by rebooting DF700

Item Description
Contents DF700 could not link up between the port of DF700 and the instrument with fixed speed setting.
The problem may occur when all the following conditions are met:
A) DF700 with 4G Fibre channel port
B) The link speed of the port at the destination instrument is setting at the fixed speed
C) Power off/on, Reboot or Online micro program exchange

9) Auto negotiation failure when FC cable is inserted in the port after removed from the port
Item Description
Contents The port speed may change from 2Gbps to 1Gbps when inserting FC cable in the port after prior
removal from the port, due to a microcode bug in link up logic of auto negotiation, function.
The problem may occur when all the following conditions are met:
A) Direct connection between HOST and DF700
B) Host environment: AIX with FC6228, FC6239 HBA
C) FC cable is inserted again after prior removal

HDS Confidential 3 of 7
10) When Subsystem boots, it does not detect failure in spite of ENC cables connected abnormally
Item Description
Contents When the Subsystem boots it does not detect the failure. The subsystem comes Ready despite
ENC cables being connected abnormally (IN and OUT are reversed). This is due to a microcode
bug in the ENC cable (IN/OUT check) during the Subsystem booting.

11) Subsystem does not come ready with "Unit which cannot be managed (Unit-XX)"
Item Description
Contents Subsystem does not come ready with "Unit which can not be managed (Unit-XX)" in following
order of connection.
RKH (DF700H) – RKA (FC) – RKA (FC) – RKAT (SATA)
Due to a microcode bug in the order of failure detection during the Subsystem booting.
The problem may occur with all the following conditions met.
A) The timing of Subsystem booting
B) DF700H
C) The last additional UNIT is RKAAT (SATA)
D) HDU (HDU-14) which is in RKA(FC) connected next to the basic UNIT

12) Execute pairresync during Quick Format for PVOL after Shadow Image I/O Switch
Item Description
Contents With Shadow Image I/O Switch Mode enabled in Shadow Image, and when the HDD that
compose PVOL are blockaded, the correct maintenance procedure is below. In the state of PSUE
(SVOL Switch), Quick Format for the PVOL is possible. No check of paireresync done for
PVOL executing Quick Format
Maintenance Procedure:
A) Exchange of the obstacle HDD
B) Execute Quick Format for the PVOL
C) Wait for completion of Quick Format
D) Execute pairresync after the (C)
In the case of (C), when pairresync executed before Quick Format is completed, the pairresync
executed incorrectly. As a result, a host may not recognize PVOL.
With all of the following conditions met, the phenomenon may occur:
A) Shadow Image is enabled
B) “Shadow Image I/O Switch Mode” enabled
C) The HDD that composes PVOL are blockaded
D) Execute Quick Format after the obstacle HDD exchange
E) Execute pairresync during Quick Format for the PVOL

13) Countermeasure for performance down in high-multiplex commands in ISCSI

Item Description
Contents The performance down for Write or Read/Write may occur in high-multiplex commands in
ISCSI. The problem may occur with all the following conditions met.
A) ISCSI
B) Many hosts (more than 40 hosts)
C) Command multiplex number is more than 512 from all hosts combined

HDS Confidential 4 of 7
14) Subsystem down with "Backend down [Cable ERR]"
Item Description
Contents Subsystem down may occur by Loop obstacle during the Subsystem booting. Problem may
occur with all the following conditions met.
A) The timing of Subsystem booting
B) The last of additional UNIT is RKAT(SATA) and connected on Path0 or Path2
C) There is Loop obstacle in CTL#0

15) Subsystem down with "Micro program error [SNR]"

Item Description
Contents Subsystem down with "Micro program error [SNR]" may occur after the Snap Shot restore is
completed.
Conditions:
A) The problem may occur when True Copy pair status changes during executing the restore of
Snap Shot
B) The problem may occur when 'pairsplit -S' executes to the pair executing Snap Shot restore
Microcode is improved to resolve this issue.

16) Subsystem not down despite of both ENC blocked

Item Description
Contents ENC#0 blocked because of removing ENC#0. ENC#1 is blocked because SES command cannot
be read. However, the subsystem is not down despite of both ENC blocked. This is due to a
microcode bug in backend diagnosis.
The problem may occur with one of the following conditions met.
A) ENC#0 blockade
B) ENC#1 blockade because of one failure except Loop failure

17) Controller blocked with DCTL failure in ISCSI

Item Description
Contents DCTL failure occurs due to the microcode bug in the ISCSI DCTL transaction.
The problem may occur with one of following conditions met.
A) ISCSI I/F
B) When the controller receives host I/O in both ports, DCTL failure occurs in Port#1 and there
is no DCTL failure in Port#0

18) Countermeasure for performance down of Random-Write

Item Description
Contents Performance of random-Write in RAID5 is down extremely temporarily. This occurs on not only
1LU (1RG), but also 2, 4, 8, 12LU. This is due to a microcode bug in the cache transaction.
The problem may occur with all the following conditions met.
A) RAID5
B) Random Write I/O
C) Large cache size

19) Illegal setting condition for directory change

Item Description
Contents Directory change does not work temporarily. Performance of I/O may be reduced.
The problem may occur with all the following conditions met.
A) Each directory assigned to each controller
B) Copies of each controller complete at the same time while both controllers execute True
Copy

HDS Confidential 5 of 7
20) System down during Snapshot restores
Item Description
Contents Restore of Snapshot can be execute by "pairresync -restore" command. System down (both
controllers failure) might occur just before finishing restore of Snapshot. Due to a microcode
bug, the end process of Snapshot restore takes more than 10 seconds in proportion to LU
capacity. DF controller failure may occur for watchdog timeout (10 seconds) at the end process
of Snapshot restore.
With all of the following conditions met, the phenomenon may occur.
A) Snapshot license is available
B) LU size of Snapshot pair is larger than 100GB
C) Executing Snapshot restore (pairresync -restore)
D) No LAN access is during the end process of Snapshot restore
(Neither Management LAN nor Maintenance LAN is accessed)
It is so that the timer of watchdog timeout is cleared by LAN access

21) Performance down in high-load heat run with low line

Item Description
Contents Performance down may occur in ISCSI high-load heat run with low line (100Mbps).
The problem may occur with all the following conditions met.
A) ISCSI I/F
B) low line(less than 100Mbps)
C) high-load, high-multiplex (more than 1280 command/port)

22) Snapshot V-VOL not deleted by Storage Navigator Modular

Item Description
Contents The problem may occur with one of following conditions met on both CTL at the same time.
A) Snapshot status becomes 'PSUE'
B) Internal job for removing V-VOL are completed

23) Changing the way of outbound check in ISCSI

Item Description
Contents ISCSI outbound EDC check hung. (countermeasure for Qlogic errata)
The problem may occur with all the following conditions met.
A) ISCSI I/F
B) Outbound Write transfer
C) Using the EDC check function

24) LA error detected in ISCSI Read transfer

Item Description
Contents LA error detected in ISCSI Read transfer while Subsystem carry the 8GB cache
The problem may occur when all following conditions are met.
A) DF700H, 8GB cache
B) ISCSI
C) Re-transfer the data of Read command

25) DF does not respond with raw I/O in NIC connection of RedhatLinux4.0U1
Item Description
Contents ISP4022 firmware bug in Linux raw I/O
The problem may occur with all the following conditions met.
A) ISCSI
B) RAW I/O from Linux
C) High-load I/O from multiple hosts

HDS Confidential 6 of 7
26) CTL reboots, (include online micro program exchange) host connection path does not recover.
Item Description
Contents When self-test program runs at port A, AIX HBA (FC6239) of direct connecting port B of
DF700 hung.
The problem may occur with all the following conditions met.
A) DF700 XS/S + 4G FC I/F
B) Direct connection from AIX+HBA(FC6239) to port B of DF700
C) When the CTL reboots (include online micro program exchange)

27) Commands timeout occurs with Snapshot function

Item Description
Contents When the Snapshot function is used a host I/O time out occurs and next I/Os is rejected.
The problem may occur with all the following conditions met.
A) Host issues Write commands to Snapshot P-VOL
B) It fails in any Snap Shot-Copy (from P-VOL to Pool) processing for CTL failure etc
C) Execution of the other Snap Shot pair operation

28) DF700 --- FC Switch ---DF700

Item Description
Contents With all the following conditions met, the problem may occur.
A) In the True Copy configuration via FC switch
B) zone configuration not configured
C) when one subsystem reboot

HDS Confidential 7 of 7