[
https://issues.apache.org/jira/browse/DERBY-4701?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Kristian Waagan updated DERBY-4701:
-----------------------------------
Attachment: hang-stack-trace.txt
Attached a sample stack trace where the ping-operation hangs. It is not clear
to me why this happens, it looks a bit like a "half-baked" shutdown.
In addition to the proposed timeout fix, we should identify the reason why this
happens [some times] when running the Derby tests.
> Set socket timeout for the socket used to communicate with the server in
> NetworkServerControl(Impl)
> ---------------------------------------------------------------------------------------------------
>
> Key: DERBY-4701
> URL: https://issues.apache.org/jira/browse/DERBY-4701
> Project: Derby
> Issue Type: Improvement
> Components: Miscellaneous, Network Server
> Affects Versions: 10.7.0.0
> Reporter: Kristian Waagan
> Attachments: hang-stack-trace.txt
>
>
> The commands in NetworkServerControl, for instance ping, are invoked over a
> socket without a timeout set. This may lead to a hang if the server ends up
> in a confused state (I have observed this when running suites.All in
> parallel).
> For this situation to arise, the server socket must be open and the server
> fail after the connection has been established, but before the reply has been
> sent to the client.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
