[
https://issues.apache.org/jira/browse/SOLR-4909?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Michael Garski updated SOLR-4909:
---------------------------------
Attachment: SOLR-4909_fix.patch
Attached a fix that corrects the issue. Here is an overview of the fix:
Simply having the slave not have a writer instance is not feasible as then old
commit points would continue to pile up on the slave, eventually exhausting
disk space. The writer is required to delete the old commit points via the
deletion policy, and the only way for the writer to be aware of the new commit
points retrieved from the master is to open a new writer, and if the reader is
reopened with a different writer instance the segment-level caches are lost.
To change this behavior the reader is disconnected from the writer when
explicitly configured to do so with a new parameter in the indexConfig section
of the solr config named createReaderFromWriter which defaults to true to make
the current behavior the default. If the value is explicitly set to false,
which would normally only be done in a read-only slave, the reader is always
initialized and re-opened from a directory instance and not a writer instance.
There is logic in SolrCore.openNewSearcher to open a new reader rather than
re-open should the underlying directory instance in the current reader not
match that of the new index writer as that means that a full copy of the index
was downloaded into a new directory, as would happen during replication if the
slave's version was ahead of the master's.
The patch was created on the lucene_solr_4_3 branch with all tests passing & I
can create versions for other branches if needed.
> Solr and IndexReader Re-opening on Replication Slave
> ----------------------------------------------------
>
> Key: SOLR-4909
> URL: https://issues.apache.org/jira/browse/SOLR-4909
> Project: Solr
> Issue Type: Improvement
> Components: replication (java), search
> Affects Versions: 4.3
> Reporter: Michael Garski
> Fix For: 5.0, 4.4
>
> Attachments: SOLR-4909-demo.patch, SOLR-4909_fix.patch
>
>
> I've been experimenting with caching filter data per segment in Solr using a
> CachingWrapperFilter & FilteredQuery within a custom query parser (as
> suggested by [[email protected]] in SOLR-3763) and encountered situations
> where the value of getCoreCacheKey() on the AtomicReader for each segment can
> change for a given segment on disk when the searcher is reopened. As
> CachingWrapperFilter uses the value of the segment's getCoreCacheKey() as the
> key in the cache, there are situations where the data cached on that segment
> is not reused when the segment on disk is still part of the index. This
> affects the Lucene field cache and field value caches as well as they are
> cached per segment.
> When Solr first starts it opens the searcher's underlying DirectoryReader in
> StandardIndexReaderFactory.newReader by calling
> DirectoryReader.open(indexDir, termInfosIndexDivisor), and the reader is
> subsequently reopened in SolrCore.openNewSearcher by calling
> DirectoryReader.openIfChanged(currentReader, writer.get(), true). The act of
> reopening the reader with the writer when it was first opened without a
> writer results in the value of getCoreCacheKey() changing on each of the
> segments even though some of the segments have not changed. Depending on the
> role of the Solr server, this has different effects:
> * On a SolrCloud node or free-standing index and search server the segment
> cache is invalidated during the first DirectoryReader reopen - subsequent
> reopens use the same IndexWriter instance and as such the value of
> getCoreCacheKey() on each segment does not change so the cache is retained.
> * For a master-slave replication set up the segment cache invalidation occurs
> on the slave during every replication as the index is reopened using a new
> IndexWriter instance which results in the value of getCoreCacheKey() changing
> on each segment when the DirectoryReader is reopened using a different
> IndexWriter instance.
> I can think of a few approaches to alter the re-opening behavior to allow
> reuse of segment level caches in both cases, and I'd like to get some input
> on other ideas before digging in:
> * To change the cloud node/standalone first commit issue it might be possible
> to create the UpdateHandler and IndexWriter before the DirectoryReader, and
> use the writer to open the reader. There is a comment in the SolrCore
> constructor by [[email protected]] that the searcher should be opened before
> the update handler so that may not be an acceptable approach.
> * To change the behavior of a slave in a replication set up, one solution
> would be to not open a writer from the SnapPuller when the new index is
> retrieved if the core is enabled as a slave only. The writer is needed on a
> server configured as a master & slave that is functioning as a replication
> repeater so downstream slaves can see the changes in the index and retrieve
> them.
> I'll attach a unit test that demonstrates the behavior of reopening the
> DirectoryReader and it's effects on the value of getCoreCacheKey. My
> assumption is that the behavior of Lucene during the various reader reopen
> operations is correct and that the changes are necessary on the Solr side of
> things.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
