Check configuration before running any benchmarks

Before doing any heavy-lifting, check existence of directories,
binaries and some other sanity checks.

3 files changed, 72 insertions, 2 deletions

diff --git a/client/benchmarks/pgbench.py b/client/benchmarks/pgbench.py
index eaa8402..2ec38d5 100644
--- a/client/benchmarks/pgbench.py
+++ b/client/benchmarks/pgbench.py
@@ -1,5 +1,6 @@
 import math
 import os
+import os.path
 import re
 import time
 
@@ -155,6 +156,35 @@ class PgBench(object):
 		return '\n'.join(o)
 
 
+	def check_config(self):
+		'check pgbench configuration (existence of binaries etc.)'
+
+		issues = []
+
+		if not os.path.isdir(self._bin):
+			issues.append("bin_dir='%s' does not exist" % (self._bin,))
+		elif not os.path.exists('%s/pgbench' % (self._bin,)):
+			issues.append("pgbench not found in bin_dir='%s'" % (self._bin,))
+		elif not os.path.exists('%s/createdb' % (self._bin,)):
+			issues.append("createdb not found in bin_dir='%s'" % (self._bin,))
+		elif not os.path.exists('%s/dropdb' % (self._bin,)):
+			issues.append("dropdb not found in bin_dir='%s'" % (self._bin,))
+		elif not os.path.exists('%s/psql' % (self._bin,)):
+			issues.append("psql not found in bin_dir='%s'" % (self._bin,))
+
+		if type(self._duration) is not int:
+			issues.append("duration (%s) needs to be an integer" % (self._duration,))
+		elif not self._duration >= 1:
+			issues.append("duration (%s) needs to be >= 1" % (self._duration,))
+
+		if type(self._runs) is not int:
+			issues.append("runs (%s) needs to be an integer" % (self._duration,))
+		elif not self._runs >= 1:
+			issues.append("runs (%s) needs to be >= 1" % (self._runs,))
+
+		return issues
+
+
 	def _run(self, duration, nclients=1, njobs=1, read_only=False, aggregate=True):
 		'run pgbench on the database (either a warmup or actual benchmark run)'
 
diff --git a/client/benchmarks/runner.py b/client/benchmarks/runner.py
index 3a891dc..b6ef293 100644
--- a/client/benchmarks/runner.py
+++ b/client/benchmarks/runner.py
@@ -32,6 +32,38 @@ class BenchmarkRunner(object):
 		self._configs.update({config_name : {'benchmark' : benchmark_name, 'config' : kwargs, 'postgres' : postgres_config}})
 
 
+	def _check_config(self, config_name):
+		''
+
+		log("checking benchmark configuration '%s'" % (config_name,))
+
+		# construct the benchmark class for the given config name
+		config = self._configs[config_name]
+		bench = self._benchmarks[config['benchmark']]
+
+		# expand the attribute names
+		bench = bench(**config['config'])
+
+		# run the tests
+		return bench.check_config()
+
+
+	def check(self):
+		'check configurations for all benchmarks'
+
+		issues = {}
+
+		if os.path.exists(self._output):
+			issues['global'] = ["output directory '%s' already exists" % (self._output,)]
+
+		for config_name in self._configs:
+			t = self._check_config(config_name)
+			if t:
+				issues[config_name] = t
+
+		return issues
+
+
 	def _run_config(self, config_name):
 		''
 
@@ -75,7 +107,6 @@ class BenchmarkRunner(object):
 	def run(self):
 		'run all the configured benchmarks'
 
-		# FIXME check that the directory does not exist
 		os.mkdir(self._output)
 
 		for config_name in self._configs:
diff --git a/client/perffarm-client.py b/client/perffarm-client.py
index 1d75a10..5826d33 100755
--- a/client/perffarm-client.py
+++ b/client/perffarm-client.py
@@ -51,4 +51,13 @@ if __name__ == '__main__':
 								postgres_config = POSTGRES_CONFIG,
 								**PGBENCH_CONFIG)
 
-		runner.run()
+		# check configuration and report all issues
+		issues = runner.check()
+
+		if issues:
+			# print the issues
+			for k in issues:
+				for v in issues[k]:
+					print k, ':', v
+		else:
+			runner.run()