Please do not kill Solr with "kill -9". Sure it stops the process but
it's A Bad Practice exactly because it dis-allows graceful shutdowns,
Solr may be forcefully killed when in the middle of operations.
You are not, by chance, specifying an async ID are you? If so, the
command will return immediately and you have to check the id via
REQUESTSTATUS to know when the command is completed.
Finally, I'd look at the Solr logs to be sure the command reached the
Solr instance in question and completed successfully. My bet: the
command didn't go where and/or do what you expected.
Best,
Erick
On Wed, May 3, 2017 at 4:21 PM, Shawn Heisey <apa...@elyograg.org> wrote:
> On 5/3/2017 1:52 PM, Shashank Pedamallu wrote:
>> I unload a core from Solr using the CoreAdmin page on UI or send unload
>> request via the api, it works perfectly fine. In my code, I’m trying to
>> unload a core using the call:
>>
>> CoreAdminRequest.unloadCore(<coreName>, <solrClient>);
>>
>> I get following response:
>>
>> {responseHeader={status=0,QTime=0}}
>>
>> At this step, I had put a debug point and Solr UI still shows the core that
>> I have unloaded. Also, I killed the solr process using kill –9 command and
>> restarted Solr. The core is still present in Solr. When does it get unloaded
>> then?
>
> Can you show how the SolrClient is created? If it's HttpSolrClient,
> then include the URL that's used to create the client. Change the
> hostname if you need to in order to disguise anything that should remain
> hidden.
>
> Ideally, I'd like to see all of the SolrJ code, and the log entry in
> solr.log from the unload call.
>
> Thanks,
> Shawn
>
