AIX Client filesystem backup fails with Error Status 10 "allocation failed"

Problem

The symptoms described here appear to be in isolation and only impact AIX Operating Systems when configured as Netbackup clients.

With this particular instance, the client policy type is set to Standard (0) backups where the All Local Drives backup would continually fail with a Netbackup Status code 10 "allocation failed". The backup would persistently fail against one particular directory which had in excess of 7,000,000 small files. 

Error Message

Netbackup Media Server.

/usr/openv/netbackup/logs/bpbrm

Critical bpbrm (pid=21172) from client womble: FTL - dir_list_rec malloc failed. Errno = 12:.

/usr/openv/netbackup/logs/bptm

Error bptm (pid=27348) media manager terminated by parent process.

Netbackup Client.

/usr/openv/netbackup/logs/bpbkar

Info bpbkar (pid=0) done. status: 10: allocation failed.

Cause

In this instance the root case is identified to be around the current ulimit settings set within the Operating System, where the data segment size is set to a lower value which is preventing the backup to complete successfully.

Also a common cause for a client backup failing with a Exit Status 10 allocation failed is insufficient system resources, so please ensure that the affected machines physical hardware has adequate resources such as physical, virtual memory and swap space to handle this type of operation.

Solution

This issue is resolved by permanently setting the Operating System ulimit value to "unlimited" to allow for enough memory to be allocated.

The Operating System ulimit value can be set with the following command: 

ulimit –d unlimited

Applies To

The reported behavior is only applicable to AIX operating system versions which are supported to be a Netbackup client.