Issue with File Scan Objects and WOB Triggers using AIX Agent

Products

• ESP Workload Automation

Issue/Introduction

Greetings - 

We're experiencing an issue with one of our AIX agents primarily used for file scans at both the workload object level and event level (WOB triggers). File scan objects periodically go into a SUBDELAY (transmitter busy) status and the agent itself experiences a connection error. In addition, we also have WOB triggers and file trigger workload objects intermittently failing (directory does not exist and/or 'java heap space' message). 

There was a recent migration from Netapp to Cohesity on this server that is also being investigated. Is there a known compatibility issue between Cohesity and ESP agents? Is there a configuration change we can make to this agent to help alleviate the issues we are experiencing? Thanks.

 

**Filesystem mount options

/aix_ap4_dw_in  cohesityp  /datawhse/inbound      nfs   --      bg,hard,intr,sec=sys,dio,llock,proto=tcp,vers=3,rsize=524288,wsize=524288 yes  no

 

 

Environment

Release : 12.0

Component : CA Workload Automation System Agent

Resolution

Cohesity support had me set the following setting which we call gflags on the Cohesity Cluster itself: 

 

gflag  bridge_nfs_enable_readdir_state_cache=false.

 

And set the AIX NFS mount options on the NFS mounted filesystems with the following options.
bg,hard,intr,sec=sys,llock,proto=tcp,vers=3,rsize=524288,wsize=524288,timeo=10000,rw