gLite 3.1
glite-CREAM - Update to version 3.1.3-0
|
| Date |
16.08.2008 |
| Priority |
Normal |
Description
LB update
glite.lb.client-interface
- install *AttrNames.pl (to be used in lb.client)
- added ui_host to jobstatus
glite.lb.ws-interface
- added ui_host (via lb.build)
- source element of genericFault is now optional
glite.lb.client
- align attr_names[] (used in getAttrName methods) in Event and JobStatus to definitions in *.h from lb.client-interface
- do the previous quick fix thoroughly -- don't smash name_with_underscores and nameWithUppercaseLetters
- added ui_host to jobstat
- store purge files (to be uploaded to JP) in GLITE_LOCATION_VAR rahter then GLITE_TMP
- use -1 as an invalid socken number instead of 0
- handle errors generated by gss inquire credentials correctly (#37084)
- correctly release unused credentials (#37069) and reuse connection
- handle gss errors whene receiving notifications (#35118)
- fix producer connection management (re-establishing connections) (#37074)
glite.lb.common
- rebuild wrt. changes in lb.build (ui_host)
- changed unused socket number from 0 to -1
glite.lb.logger
- startup script: use grep with ^ rather than fgrep
- fixed sending empty events
- do not syslog "error reading server reply" unnecessarily
glite.lb.server
- store dumps in GLITE_LOCATION_VAR rather than /tmp
- don't waste memory while reindexing
- handle subjob embryos in reindexing
- unify error handling for store and query
- don't clobber store errors with job unlocking
- don't send CollectionState from lb proxy to server
- export UI host and VOMS FQAN to RGMA
- don't coredump on requesting notifications that are denied
- don't return empty results for some non-indexed queries
- clarified soap error messages
glite.lb.proxy
- follow changes in server-1.9.0-1
glite.lb.server-bones
glite.lb.utils
- added fields GLITE_JP_LB_finalDoneStatus, GLITE_JP_LB_finalStatusDate
- local implementation of JPPS backend functions to make statistics really working
- added an offline purging script
glite-BDII
This update of the bdii fixes a critical bug which could potentially cause problems on any node type if the variables BDII_VAR_DIR or BDII_DIR are not set in the configuration file.
In addition the the attributed set_lk_max_locks 2000 has been set in the DB_CONFIG file which is a partial solution to Bug #42727.
Removal of glite-info-plugin-fcr solving bug #40649
Please also have a look at the list of known issues.
This update fixes various bugs. For the full list of bugs, please see list below.
Fixed bugs
| Number | Description |
|
#21575 |
add VOMS attributes to the L&B rgma feed |
|
#21578 |
add UI to the L&B rgma feed |
|
#28770 |
LB server syslogs client-side errors generally irrelevant to admin |
|
#30203 |
LBproxy sends CollectionState events to LB server |
|
#31829 |
glite-yaim-core does use SL4 only sed options |
|
#31944 |
unindexed LB queries return no results |
|
#33397 |
glite-lb-locallogger script bug |
|
#33777 |
glite-lb-bkserverd must ignore SIGXFSZ |
|
#35118 |
LB notifications: garbage comming to client |
|
#35490 |
glite-lb-bkindex uses too much memory |
|
#35623 |
LB server fails to report back error encountered while storing LB event |
|
#35632 |
glite-lb-bkindex may abort on embryonic subjobs |
|
#37069 |
memory leaks in LB client library |
|
#37074 |
LB client library does not handle closing of kept-alive logd connection correctly |
|
#37084 |
LB client library can cause segfault when credential expires |
|
#42799 |
bdii: varibles not check before running chown |
Updated rpms
The RPMs can be updated using yum via
Service reconfiguration after update
Not needed.
Service restart after update
Not needed.
How to apply the fix
- Update the RPMs (see above)
- Update configuration (see above)
- Restart the service if necessary (see above)
|
|
