[jira] [Commented] (SOLR-12998) Race condition between requery recovery and the core load completing

["Nick Shanin (Jira)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/SOLR-12998?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18122095#comment-18122095 ] 

Nick Shanin commented on SOLR-12998:
------------------------------------

🤖 AI text below 🤖 (posted on behalf of Nick Shanin)

I opened a PR for this: https://github.com/apache/solr/pull/5010

REQUESTRECOVERY now returns a 503 instead of a 400 when the named core is still loading, and SyncStrategy retries that exact response up to two more times; other 503 responses are not retried. Tests cover the 503/400 split and the retry behaviour, including no retry for a different 503 and consistent Error propagation.

> Race condition between requery recovery and the core load completing
> --------------------------------------------------------------------
>
>                 Key: SOLR-12998
>                 URL: https://issues.apache.org/jira/browse/SOLR-12998
>             Project: Solr
>          Issue Type: Bug
>            Reporter: Varun Thacker
>            Priority: Major
>              Labels: pull-request-available
>          Time Spent: 10m
>  Remaining Estimate: 0h
>
> I started Solr ( 7_6 branch ) using {{bin/solr start -e cloud -noprompt}} and indexed 25M documents which amounts to roughly 1GB per replica in raw disk space.
>  
> Whenever I restart the cluster, node2 ( 7574 ) get's these errors in the logs which seems to indicate that request recovery is trying to find a core which isn't completely loaded yet.
>  
> {code:java}
> 2018-11-18 21:06:09.845 INFO  (coreLoadExecutor-9-thread-2-processing-n:192.168.0.7:7574_solr) [c:gettingstarted s:shard2 r:core_node8 x:gettingstarted_shard2_replica_n6] o.a.s.c.SolrCore [[gettingstarted_shard2_replica_n6] ] Opening new SolrCore at [/Users/varunthacker/apache-work/lucene-solr/solr/example/cloud/node2/solr/gettingstarted_shard2_replica_n6], dataDir=[/Users/varunthacker/apache-work/lucene-solr/solr/example/cloud/node2/solr/gettingstarted_shard2_replica_n6/data/]
> ...
> 2018-11-18 21:06:10.501 INFO  (qtp690521419-148) [   x:gettingstarted_shard2_replica_n6] o.a.s.h.a.CoreAdminOperation It has been requested that we recover: core=gettingstarted_shard2_replica_n6
> ...
> 2018-11-18 21:06:10.503 ERROR (qtp690521419-141) [   x:gettingstarted_shard1_replica_n1] o.a.s.h.RequestHandlerBase org.apache.solr.common.SolrException: Unable to locate core gettingstarted_shard1_replica_n1
>         at org.apache.solr.handler.admin.CoreAdminOperation.lambda$static$5(CoreAdminOperation.java:167)
>         at org.apache.solr.handler.admin.CoreAdminOperation.execute(CoreAdminOperation.java:360)
>         at org.apache.solr.handler.admin.CoreAdminHandler$CallInfo.call(CoreAdminHandler.java:395)
>         at org.apache.solr.handler.admin.CoreAdminHandler.handleRequestBody(CoreAdminHandler.java:180)
> ...
> 2018-11-18 21:06:10.503 INFO  (qtp690521419-148) [   x:gettingstarted_shard2_replica_n6] o.a.s.s.HttpSolrCall [admin] webapp=null path=/admin/cores params={core=gettingstarted_shard2_replica_n6&action=REQUESTRECOVERY&wt=javabin&version=2} status=400 QTime=3
> ... still the core is loading
> 2018-11-18 21:06:10.876 INFO  (coreLoadExecutor-9-thread-2-processing-n:192.168.0.7:7574_solr) [c:gettingstarted s:shard2 r:core_node8 x:gettingstarted_shard2_replica_n6] o.a.s.s.SolrIndexSearcher Opening [Searcher@2d17168a[gettingstarted_shard2_replica_n6] main]
> ... now done
> 2018-11-18 21:06:10.953 INFO  (searcherExecutor-10-thread-1-processing-n:192.168.0.7:7574_solr x:gettingstarted_shard2_replica_n6 c:gettingstarted s:shard2 r:core_node8) [c:gettingstarted s:shard2 r:core_node8 x:gettingstarted_shard2_replica_n6] o.a.s.c.SolrCore [gettingstarted_shard2_replica_n6] Registered new searcher Searcher@2d17168a[gettingstarted_shard2_replica_n6] main{ExitableDirectoryReader(UninvertingDirectoryReader(_49(7.6.0):C2416916/4913:delGen=2 _a4(7.6.0):C2474322 _71(7.6.0):C2723220 _4k(7.6.0):c111317 _d7(7.6.0):C2501101 _bt(7.6.0):c91922 _di(7.6.0):c339286 _eb(7.6.0):c163244 _ds(7.6.0):c92780 _e1(7.6.0):c337329 _em(7.6.0):c360003 _ew(7.6.0):c263533 _fg(7.6.0):c226118 _f6(7.6.0):c266132 _es(7.6.0):C2361 _f8(7.6.0):C2251 _ff(7.6.0):C1821 _fh(7.6.0):C20079 _fi(7.6.0):C35733 _fj(7.6.0):C37203 _fk(7.6.0):C34780 _fl(7.6.0):C262 _fm(7.6.0):C2148 _fn(7.6.0):C2503))}{code}



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@solr.apache.org
For additional commands, e-mail: issues-help@solr.apache.org