Automatic tests:

• The following tests of the CREAM testsuite fail:
  • Test #3: Cancellation errors due to bug #44508
  • Test #5: Submission fails due to bug #45914

Checked bugs:

• Bug #16761: Various possible execution error in BLAH when logging for DGAS FIXED
  • Tested as described here

• Bug #29679: AFS token grabbing via gssklog from job wrappers FIXED
  • Verified by the guy who submitted the bug. See here

• Bug #41215: IPv6 bug: non compliant wildcard inet address in source code HOPEFULLY FIXED
  • Verified that the Wildcard inet address INADDR_ANY has been removed from that file. No tests done on IPv6 (they will be done by SA2 as agreed)

• Bug #41216: IPv6 bug: non compliant wildcard inet address in source code HOPEFULLY FIXED
  • Verified that the Wildcard inet address INADDR_ANY has been removed from that file. No tests done on IPv6 (they will be done by SA2 as agreed)

• Bug #41219: IPv6 bug: non compliant wildcard inet address in source code HOPEFULLY FIXED
  • Verified that the Wildcard inet address INADDR_ANY has been removed from those files. No tests done on IPv6 (they will be done by SA2 as agreed)

• Bug #41221: Pv6 bug: non compliant data structure (sockaddr_in ) used HOPEFULLY FIXED
  • Verified that sockaddr_in has been removed from those files (actually there is still a struct sockaddr_in servAddr; in BPRclient.c but that variable is not used at all). No tests done on IPv6 (they will be done by SA2 as agreed)

• Bug #41223: IPv6 bug: non compliant name resolving function in source code (gethostbyname) HOPEFULLY FIXED
  • Verified that gethostbyname has been removed from those files. No tests done on IPv6 (they will be done by SA2 as agreed)

• Bug #41224: IPv6 bug: non compliant address casting function in source code HOPEFULLY FIXED
  • Verified that net_ntoa has been removed from those files. No tests done on IPv6 (they will be done by SA2 as agreed)

• Bug #41226: IPv6 bug: non compliant address casting function in source code (inet_aton) HOPEFULLY FIXED
  • Verified that inet_aton has been removed from those files. No tests done on IPv6 (they will be done by SA2 as agreed)

• Bug #42401: CREAM doesn't properly manage the JDL attribute MWVersion FIXED
  • Verified as explained here

• Bug #43830: Problems if there are > 32000 active jobs per user on a CREAM CE FIXED
  • Verified as explained here

• Bug #44888: Issue with CREAM template configuration file FIXED
  • Verified as explained here

• Bug #44889: Issue with CEMon template configuration file FIXED
  • Verified as explained here

• Bug #45253: executing /opt/glite/libexec/jobwrapper in blah FIXED
  • Checked on a WN where /opt/lcg/libexec/jobwrapper exists. It was verified that this was executed
  • Checked on a WN where both /opt/lcg/libexec/jobwrapper and /opt/glite/libexec/jobwrapper exist. It was verified that the latter was executed.

• Bug #45717: BLParserPBS should consider log lines like "unable to run job" HOPEFULLY FIXED
  • How to test this bug is described here. Unfortunately I don't know how to reproduce that problem.

• Bug #46508: ban-plugin should be initialized in cream-config.xml FIXED
  • verify that the authorization section of the cream-config.xml file contains the ban plugin.

• Bug #46722: Issue with creation of sandbox dir in CREAM FIXED
  • Verified that the CREAM conf file (cream-config.xml) CREAM_CREATE_SANDBOX_BIN_PATH refers to the new executable /opt/glite/bin/glite-cream-createsandboxdir

• Bug #47461: The jobPurge command must be executed with low priority HOPEFULLY FIXED
  • To check the priority of the commands, the DB should be checked. But when a job is purged, the entry for that job is removed from the DB. So it should be checked when there are many commands in the queue, before the purge is actually executed

• Bug #48083: CREAM - unhelpful error for glexec problem FIXED
  • Verified as explained here

• Bug #48144: Problems with purge in CREAM when the mapped group name is different than the VO name FIXED
  • Verified as explained here

• Bug #48263: cannot store the delegation proxy locally HOPEFULLY FIXED
  • run cream-test-monitored-submit command with high values of the options -n, -m and -C

• Bug #48302: cannot store the delegation proxy locally FIXED
  • Verified as described here.

• Bug #48482: cannot store the delegation proxy locally FIXED
  • Check here to understand how to test the fix.
  • The problem was happening at CNAF. After having applied the fix, it didn't appear anymore

• Bug #49284: Wrong format for BlackListServicePDP config file FIXED
  • insert the DN of the user in quotes in the lcas config file (/opt/glite/etc/lcas/ban_users.db)
  • try any command and verify that the user is not allowed
  • remove the DN
  • try any command and verify the the user is allowed again
  • this bug is related to #46508 (see above)

• Bug #49514: Issue with requirements forwarding and 'member' function FIXED
  • Tested as described here

• Bug #49557: BLParserLSF problem with lsid at startup FIXED
  • Verified as described here.

• Bug #49932: Maximum number of active db connections exceeded NOT REPRODUCIBLE
  • lower the value of the attribute maxActive in the file /usr/share/tomcat5/conf/Catalina/localhost/ce-cream.xml
  • run the testsuite with a number of submission higher than the attribute value ( options -n -m -C)
  • This test fails due to bug #48801

• Bug #50163: delegationProxyAdd raises a Nullpointerexception FIXED
  • delegate a new proxy using the command glite-ce-delegate-proxy
  • verify that in the service log does not appear the error "cannot store the delegation proxy locally"

-- PaoloAndreetto - 24 Apr 2009