Solaris Agents 12.2.x / 12.3.x crash upon startup on hostname/ip resolution


Article ID: 192841


Updated On:


CA Automic Workload Automation - Automation Engine CA Automic Workload Automation - Automation Engine


Recent Solaris Agents (tested 12.2.1hf3 / 12.2.5 / 12.3.2) crash generating a core upon startup while resolving either the current hostname IP or the Automation Engine IPs.

Error when launching the Agent:

[1]+  Segmentation Fault      nohup ./ucxju64


The core shows that the crash occurs on the OS APIs gethostbyname() and getaddrinfo() 


The Systems where the issue occurs are:

OS: SunOS xxxx 5.10 Generic_150400-17 sun4v sparc SUNW,SPARC-Enterprise-T5120

Network configuration: Solaris Cluster is configured with the following line in /etc/nsswitch.conf:

hosts: cluster files dns


Old Agents on version 11.2.2 can start and resolve the ip addresses correctly with the same port / binaddr and cp settings connecting to the same Automation Engine.

[1]+  Segmentation Fault      nohup ./ucxju64


Unknown, most likely due to the /etc/nsswitch.conf cluster configuration


Release : 12.2.x and 12.3.x


Subcomponent: Agent for Unix ( Solaris Sparc64)


Root cause could not be investigated as the problematic Solaris 10 Servers were to be commissioned in a few months.

The workaround consists on using Solaris Agents 11.2.x or 12.0.x instead.


Additional Information

If you experience this issue, please open a ticket with Support to try and find a Solution for Agents 12.3.x