[
https://issues.apache.org/jira/browse/CLOUDSTACK-4490?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13852611#comment-13852611
]
Sowmya Krishnan commented on CLOUDSTACK-4490:
---------------------------------------------
This is an intermittent issue. The likely cause is as follows:
When we login to NS, we are by default, logged into the CLI. To exit from the
CLI, we can either do a “logout” or “exit” explicitly. We can try implementing
this in the automation scripts too by calling an explicit logout or exit after
we are done with validation. I am not yet sure if this will raise an error when
ssh.close() is invoked. We can test it out after fixing.
> [Automation] ssh.close not happening leading to potential failures in
> Netscaler test scripts
> --------------------------------------------------------------------------------------------
>
> Key: CLOUDSTACK-4490
> URL: https://issues.apache.org/jira/browse/CLOUDSTACK-4490
> Project: CloudStack
> Issue Type: Test
> Security Level: Public(Anyone can view this level - this is the
> default.)
> Components: Automation
> Affects Versions: 4.2.0
> Environment: 4.2, Netscaler as public LB provider
> Reporter: Sowmya Krishnan
> Fix For: 4.3.0
>
>
> ssh connections through remoteSSHClient.py not closing connections causing
> potential failures in Netscaler scripts. This is generally not a problem with
> VMs since we cleanup up the account after the script ends, but with ssh to
> external devices like Netscaler, we end up with too many open sessions
> causing NS to refuse more connections.
> We end up with the following error in NS:
> Error: Connection limit to CFE exceeded
> (This happens to be a known issue with certain versions of NS)
> In any case, I think we should either fix remoteSSHClient or add an explicit
> close in the test scripts.
--
This message was sent by Atlassian JIRA
(v6.1.4#6159)
