Summary
Block backups and Catalog Condense jobs fail after upgrading a UNIX or Linux master server to DPX 4.4.2.
Symptoms
The job log attributes the failures to an inability to connect to the Snapvault Handler (SVH) Backup Express module with an error code of 12104.
xxx.xxx.xxx.58 12/6/2007 2:49:48pm SNBSCH5607J Starting job BE_CONDENSE (start job with sched_id 6974188)
xxx.xxx.xxx.58 12/6/2007 2:59:50pm SNBRTM1101E CM API function: CM_AP_CONNECT_PLIM return code: 12104 on socket: 65535
xxx.xxx.xxx.58 12/6/2007 2:59:50pm SNBRTM5693J connecting to sssvh at test_node failed (rc = 12104)
xxx.xxx.xxx.58 12/6/2007 2:59:50pm SNBRTM5769E Task CONDENSE (1196974188) can not connect to snapvault job handler (rc = 12104)
Resolution
Check Svhstdlog, located on the master server ($SSPRODIR\logs), it may display the following error message:
SVH starting ...
# There is insufficient memory for the Java Runtime Environment to continue.
# pthread_getattr_np
# An error report file with more information is saved as:
# /opt/DPX/bin/hs_err_pid416.log [thread 140681667639040 also had an error]
SVH exited
This means that there is not enough physical Memory on the DPX Master Server for Snapvault Handler to start. The minimum requirement for Catalogic DPX Master Server is 4GB of memory. Recommended physical memory is 8GB.
Catalogic has seen this failure when customer’s physical memory on Catalogic DPX Master Server was set to only 1GB of memory. Ensure your Catalogic DPX Master Server has sufficient amount of free memory.