Test bunch #1: requests:
#1.1:
#1.1.1: submit a simple "Hello, world" job with trivial ISB and OSB.
#1.1.2: submit a collection made of a bunch of such jobs, say 10.
check all option, especially json compliance
test submission to CREAM
test submission to ICE
glite-wms-job-cancel
#1.2.1: glite-wms-job-cancel for single job
#1.2.2: glite-wms-job-cancel for collection job parent and nodes. pending nodes should be at various mixed states (submitted, ready, scheduled, running)
check all option, especially json compliance
glite-wms-job-info
glite-wms-job-logging-info
check that
ReallyRunning event is present in
LogMonitor
glite-wms-job-perusal
glite-wms-job-submit
glite-wms-job-delegate-proxy
glite-wms-job-list-match
check all option, especially json compliance
glite-wms-job-output
glite-wms-job-status
check all option, especially json compliance
glite-wms-proxy-sign
Test bunch #2: job types:
-collection
use both: node = [a=b;c=d;...] and node = [file = "..."]
-dag
DAG1: dag with no dependencies
DAG2: dag with terminal failing node
DAG3: dag with non-terminal failing node
max_running_nodes,NodesCollocation
-parametric
-MPI
Test bunch #3: security features:
delegation: automatic, explicit
voms proxy: old-style, --rfc, no attributes, one attribute, more attributes
authorization: check how gacl file is generated and match by dn and fqan.
proxy renewal: myproxyserver="myproxy.cnaf.infn.it" check with both ICE and jc
sandbox mapping with gridftp and lcas, lcmaps. check that no mix up is done when using delegations from the same user with different certificates/attributes
Test bunch #4: server features:
check that init scripts work with all their options (start, stop, graceful, etc.)
check limiter kicking in
check wm recovery
check stale gridftp processes (#53700)
stop the wm with pending jobs and check that at restart they are properly restored from their last state
create a dump of the ism, adding a request on file like [command="ism_dump";]
check that proxycache cron purges expires delegated proxies.
check that lb proxy does NOT purge done jobs before one week
check that wms purger works properly
check
ExpiryPeriod weith jobs that do not specify it in their jdl
check
MatchRetryPeriod
check
MaxOutputSandboxSize
check ice configuration
check wmproxy in static and dynamic mode (now with fcgid)
check WMS+LB co-hosting, checklbproxy=false/true
job-list-match: requirements =true, requirements=false
job submit
1) ISB
2) ISB with one empty file
3) zipped ISB
4) OSB
5) ISB+OSB
after job is complete:
- check that proxy is deregistered
- sandbox is purged
- check that lb proxy did NOT purge done the job before one week
Test bunch #5: JDL features:
-nodescollocation
mm with data
-inputdata lfn:, guid:, lds and query
-DataRequirements
-DataCatalog,
DataCatalogType,
DataAccessProtocol,
OutputSE
-Gang-matching
-Resubmission:
- deep
- RetryCount
- shallow
- shallowRetryCount
- requirements, rank, fuzzyrank
-
ExpiryTime (=1)
- resubmission
- zippedisb
- replanning
- Environment (Environment={"CIAO=1008"} executable= echo $CIAO
Test bunch #6: Jobwrapper:
sandbox splitting
sandbox tracking
Test bunch #7: Configuration:
metapackage, dependencies
glue and glie2.0 service information
regression tests
submit a job with expired ACs but valid certificate and check that fqan authZ doesn't let the job pass
URLs in ISB are sometimes converted to lowercase by WMproxy
Detect job status changes in ICE->CREAM with secondary FQANs
Env variables, ~ character are not correctly expanded in the WMS UI
#77284 WMProxy deleagation ID should be associated to DN + FQAN
#77282 WMS information system usage is case-sensitive
#74832 Files specified with absolute paths shouldn't be used with inputsandboxbaseuri
#74737 glite-wms-wmproxy start removes fastcgi socket !
#74221 Perusal doesn't work with nodes of a collection
#73329 ICE and CREAM can loose eventid synchronization
#73286 glite-wms-job-status and glite-wms-job-logging info wrongly see a corrupted help file
#72870 The WMS-UI does not check type mismatch for SMPGranularity,
WholeNodes and
HostNumber
#71438 proxycache purger fails on symlinked proxy cache
#70331 glite-wms-create-proxy "ambiguous redirect"
#70401 stale Condor-G jobs slowly piling up in held state
#70061 WMS hates collections with 192 nodes!
#68307 WMS-UI: syntax error for correct JDL when substr is used
#68786 "InputSandboxBaseURI" JDL attribute work only in some circumstances...
#66721 Ineffective and never removed Job cancels
#64698 jobwrapper max osb limit should be considered only if the gridftp server is the wms
#64462 high CPU usage of workload manager in gLite-WMS
#64567 job from gLite-WMS is still running after proxy expiration
#63824 LCMAPS excessive logging
#63113 WMS should allow any groups/roles by default and #50009 wmproxy.gacl person record allows anyone to pass !!!!
#59453 ICE polling needs to be improved
#59502 WM became unresponsive after St9bad_alloc
#59611 WMS UI: should transfer files even if size is 0 bytes
#59781 limit maximum sleep time in job wrapper
#56827 workload manager needs better ISM logging
#56734
ListMatch should consider also SDJ specification
#56673 WMProxy SL5: Gridsite is not backward compatible with old style proxy
#56395 WMProxy fails to enqueue jobs and they remain Submitted forever
#56330 httpd doesn't start after upgrade WMS
#56090 Correctly submitted JDL staying in SUBMITTED - error parsing classad
#56034 Matchmaking with
JobType=Normal does not take
NodeNumber into account
#55684 WMSProxy(wmproxy) gets FQAN wrong using mod_gridsite 1.7.4 Patch
#55649 WMS 3.2: wmproxy/fastcgi crashes if gcal contains invalid FQAN
#55606 glite-wms-job-listmatch is sometimes slow
#55532 WMS only accepts LSF/PBS batch system for MPI jobs when expressed as MPICH
#55452 CMS production struck by waves of "Globus error 10: data transfer to the server failed"
#54728 WMP finds FQAN inconsistency only if GROUPS are different, not ROLES
#54079 glite-wms-job-submit segmentation if requirements=();
#53733 The
StorageElements section of the
BrokerInfo files contains some repetitions
#53714 WMS PURGER SHOULD NOT directly FORCE PURGE OF jobs when its DN is not authorized on LB server
#53700 HUGE NUMBER OF GRIDFTP CONNECTION in CLOSE_WAIT STATUS
#53294 WMS 3.2 WMProxy logs are useless below level 6
#52371 [ yaim-wms ] /var/log on WMS ends up owned by glite
#52003 ICE crashes when the purger is called
#51296 glite-wms-job-info: does not recognize if a job is simple or DAG/Collection
#51295 glite-wms-job-output: fails creating the output directory if not present
#51294 glite-wms-job-submit:
ExpiryTime wrong type
#51293 glite-wms-job-status does not load LBAddresses attribute
#51292 glite-wms-job-status does not load default configuration file
#48640 glite-wms-wmproxy to support graceful command
#48598 job submission fails when first WMS tried is draining !!
#48172 wmproxy stop/restart can leave old processes behind
#48079 WMProxy has garbage in
GlueServiceStatusInfo
#48068 [wms]
GlueServiceStatusInfo content is ugly
#47404 Ambiguous error message 'System load is too high:'
#47150 glite-wms-wm script: problem when moving files from too much populated directories
#33342: separate retry policies for ISB and OSB
#36292: Not all attributes of a SA/SE coul be used in a gangmatching
When a collection is aborted the "Abort" event should be logged for the sub-nodes as well
#40982: When a collection is aborted the "Abort" event should be logged for the sub-nodes as well /2
#44599: WMS should consider
MaxTotalJobs
#45883: Optimization of resubmission
#48636: job wrapper should log events for truncated files
#49844: WMProxy does not catch signal 25
#52617: [ yaim-wms ] host{cert,key}.pem in /home/glite
#56933: WMProxy Server: gSoap needs to be built with WITH_IPV6 flag
#58878: Request for a feature allowing propagation of generic parameters from JDL to LRMs
#61557: user job is not killed when proxy expires
#70824: environment values in JDL cannot have spaces
#78030: Alternative GLITE_WMS_LOG_DESTINATION in the jobwrapper
#77004: Wrong myproxyserver string processing in ICE
#75402: Synchronization loss between real validity of proxy and exp. time saved in ICE's database
#74259: Previous matches information is not taken into account if direct submission is used