Fix Readme
Abstract
CL_HACMP (Doc Number=4832): High Impact / Highly Pervasive APARs IV74438/IV77384/IV80836/IV81777 IV74438/IV77384/IV80836/IV81777 - Cluster node halt due to Group Services client errors
A node may panic when it should not. The PANIC STRING in
errpt will show "RSCT reboot caused by critical resource
protection - Group Services".
The intended purpose of this panic is to protect the cluster
when a vital daemon becomes unresponsive. However, there are
multiple apars in which this can erroneously occur during
normal conditions:
IV74438) A long-blocking system call holds a lock blocking
the main thread of an RMC subsystem process
IV77384) Trace corruption leads to deadlock
IV80836) A blocking system call in a critical code path
IV81777) Bad HMC message causes RMC core on LPAR
It is recommended that all PowerHA users ensure they
have all these issues fixed to avoid any inappropriate node
panics.
Note that all these fixes are included in the PowerHA
SystemMirror Recommended Software Levels available at
http://ibm.biz/PowerHAFixes
Affected AIX Levels and Recommended Fixes
Minimum Level Maximum Level APAR Fixes Interim Fixes 6100-09-00
rsct.basic.rte 3.1.5.0
rsct.core.rmc 3.1.5.0 6100-09-07
rsct.basic.rte 3.2.0.7
rsct.core.rmc 3.2.0.9 IV74438
IV77384
IV80836
IV81777 Combined iFix 7100-03-00
rsct.basic.rte 3.1.5.0
rsct.core.rmc 3.1.5.0 7100-03-06
rsct.basic.rte 3.2.0.7
rsct.core.rmc 3.2.0.9 IV74438
IV77384
IV80836
IV81777 Combined iFix 7100-04-00
rsct.basic.rte 3.2.1
Content
IV74438/IV77384/IV80836/IV81777 - Cluster node halt due to Group Services client errors
A node may panic when it should not. The PANIC STRING in
errpt will show "RSCT reboot caused by critical resource
protection - Group Services".
The intended purpose of this panic is to protect the cluster
when a vital daemon becomes unresponsive. However, there are
multiple apars in which this can erroneously occur during
normal conditions:
IV74438) A long-blocking system call holds a lock blocking
the main thread of an RMC subsystem process
IV77384) Trace corruption leads to deadlock
IV80836) A blocking system call in a critical code path
IV81777) Bad HMC message causes RMC core on LPAR
It is recommended that all PowerHA users ensure they
have all these issues fixed to avoid any inappropriate node
panics.
Note that all these fixes are included in the PowerHA
SystemMirror Recommended Software Levels available at
http://ibm.biz/PowerHAFixes
Affected AIX Levels and Recommended Fixes
| Minimum Level | Maximum Level | APAR Fixes | Interim Fixes |
|---|---|---|---|
| 6100-09-00 rsct.basic.rte 3.1.5.0 rsct.core.rmc 3.1.5.0 | 6100-09-07 rsct.basic.rte 3.2.0.7 rsct.core.rmc 3.2.0.9 | IV74438 IV77384 IV80836 IV81777 | Combined iFix |
| 7100-03-00 rsct.basic.rte 3.1.5.0 rsct.core.rmc 3.1.5.0 | 7100-03-06 rsct.basic.rte 3.2.0.7 rsct.core.rmc 3.2.0.9 | IV74438 IV77384 IV80836 IV81777 | Combined iFix |
| 7100-04-00 rsct.basic.rte 3.2.1.0 rsct.core.rmc 3.2.1.0 | 7100-04-02 rsct.basic.rte 3.2.1.10 rsct.core.rmc 3.2.1.10 | IV85893 | iFix |
| 7200-00-00 rsct.basic.rte 3.2.1.0 rsct.core.rmc 3.2.1.0 | 7200-00-02 rsct.basic.rte 3.2.1.10 rsct.core.rmc 3.2.1.10 | IV85893 | iFix |
The table above describes which active levels are affected and where to obtain fixes.
Prior to the APAR Fix being available, an Interim Fix (iFix) is available for each affected level.
The available iFixes may only apply to the latest Service Packs. If a custom iFix is required please contact IBM Support.
The iFixes can be downloaded from the same location using ftp, http, or https.
| Doc number: 4832 | Published date: 20160628 |
Was this topic helpful?
Document Information
Modified date:
28 June 2016
UID
isg1SSRVPOCL_CLCL_HACMP160628-0813