Checking 'java.io.tmpdir' exists and is readable/writable

Jay Shaughnessy jshaughn at redhat.com
Tue Oct 1 21:16:13 UTC 2013


I think I'm for aborting the startup completely with messaging.  I think 
it's likely to be unintentional, and I think it's typically an 
assumption when writing the code that it will be writable.

On 10/1/2013 12:38 PM, Thomas Segismont wrote:
> Hi everyone,
>
> We had a few problems reported recently and it appeared they were all 
> related to the 'java.io.tmpdir' not existing or not being 
> readable/writable. Reports were for Windows platform, but this could 
> happen on other platforms as well (nothing prevents you from defining 
> this Java system property instead of using the default value).
>
> As we heavily rely on this directory, Larry proposed we add a check 
> and fail the server startup if needed.
>
> I think adding a check is a good idea and that a warning message (in 
> the server logs and why not in the UI) would be enough as the problem 
> does not prevent RHQ from running core services (like handling 
> inventory and measurement reports).
>
> Any thoughts?
>
> Thanks,
> Thomas
>
> _______________________________________________
> rhq-devel mailing list
> rhq-devel at lists.fedorahosted.org
> https://lists.fedorahosted.org/mailman/listinfo/rhq-devel



More information about the rhq-devel mailing list