[
https://issues.apache.org/jira/browse/OAK-8912?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17100663#comment-17100663
]
Julian Reschke edited comment on OAK-8912 at 5/6/20, 12:09 PM:
---------------------------------------------------------------
1) 1.10.2 is outdated; please use the latest release from the current
maintenance branch 1.22 (1.22.3 right now)
2) If you can reproduce this, please attach sufficient code so that we can as
well. Optimally, see the existing VersionGC tests and see why they pass (what's
the difference compared to your setup?)
3) The exception indicates that a connection wasn't properly closed; setting
the system property
"org.apache.jackrabbit.oak.plugins.document.rdb.RDBConnectionHandler.CHECKCONNECTIONONCLOSE"
to "true" might give you better diagnostics.
was (Author: reschke):
1) 1.10.2 is outdated; please use the latest release from the current
maintenance branch 1.22 (1.22.3 right now)
2) If you can reproduce this, please attach sufficient code so that we can as
well. Optimally, see the existing VersionGC tests and see why they pass (what's
the difference to you setup?)
3) The exception indicates that a connection wasn't properly closed; setting
the system property
"org.apache.jackrabbit.oak.plugins.document.rdb.RDBConnectionHandler.CHECKCONNECTIONONCLOSE"
to "true" might give you better diagnostics.
> Version garbage collector is not working if documents exceeded 100000
> ---------------------------------------------------------------------
>
> Key: OAK-8912
> URL: https://issues.apache.org/jira/browse/OAK-8912
> Project: Jackrabbit Oak
> Issue Type: Bug
> Components: documentmk
> Reporter: Ankush Nagapure
> Priority: Major
> Attachments: exception.txt
>
>
> Oak version - 1.10.2, PostgreSQL 10.7 (10.7), using driver: PostgreSQL JDBC
> Driver 42.2.2 (42.2).
> *Actual :*
> After below code run, if document collectlimit exceeded 100000, it throws
> exception attached in .txt
> {color:#0747a6}public static void runVersionGC() { {color}
> {color:#0747a6} log.info("Running garbage collection for
> DocumentNodeStore");{color}
> {color:#0747a6} try {{color}
> {color:#0747a6} final VersionGCOptions versionGCOptions =
> new VersionGCOptions();{color}
> {color:#0747a6}
> *versionGCOptions.withCollectLimit(1000000);*{color}
> {color:#0747a6}
> *documentNodeStore.getVersionGarbageCollector().setOptions(versionGCOptions);*{color}
> {color:#0747a6} log.info("versionGCOptions.collectLimit : "
> + versionGCOptions.collectLimit);{color}
> {color:#0747a6}
> documentNodeStore.getVersionGarbageCollector().gc(0, TimeUnit.DAYS);{color}
> {color:#0747a6} } catch (final DocumentStoreException e) {{color}
> {color:#0747a6} //{color}
> {color:#0747a6} }{color}
> Below is the code to create repository and get documentNodeStore object for
> version garbage collection.
> {color:#0747a6}private static Repository createRepo(final Map<String,
> String> dbDetails){color}
> {color:#0747a6} throws DataStoreException {{color}
> {color:#0747a6} try {{color}
> {color:#0747a6} final RDBOptions options ={color}
> {color:#0747a6} new
> DBOptions().tablePrefix(dbDetails.get(DB_TABLE_PREFIX)).dropTablesOnClose({color}
> {color:#0747a6} false);{color}
> {color:#0747a6} final DataSource ds ={color}
> {color:#0747a6} RDBDataSourceFactory.forJdbcUrl({color}
> {color:#0747a6} dbDetails.get("dbURL"),{color}
> {color:#0747a6} dbDetails.get("dbUser"),{color}
> {color:#0747a6} dbDetails.get("dbPassword"));{color}
> {color:#0747a6} final Properties properties =
> buildS3Properties(dbDetails);{color}
> {color:#0747a6} final S3DataStore s3DataStore =
> buildS3DataStore(properties);{color}
> {color:#0747a6} final DataStoreBlobStore dataStoreBlobStore = new
> DataStoreBlobStore(s3DataStore);{color}
> {color:#0747a6} final Whiteboard wb = new
> DefaultWhiteboard();{color}
> {color:#0747a6} bapRegistration ={color}
> {color:#0747a6}
> wb.register(BlobAccessProvider.class,(BlobAccessProvider) {color}
> {color:#0747a6}
> dataStoreBlobStore,properties);{color}
> {color:#0747a6} *documentNodeStore =*{color}
> {color:#0747a6} *new RDBDocumentNodeStoreBuilder()*{color}
> {color:#0747a6} *.setBlobStore(dataStoreBlobStore)*{color}
> {color:#0747a6} *.setBundlingDisabled(true)*{color}
> {color:#0747a6} *.setRDBConnection(ds, options)*{color}
> {color:#0747a6} *.build();*{color}
> {color:#0747a6} repository = new
> Jcr(documentNodeStore).with(wb).createRepository();{color}
> {color:#0747a6} return repository;{color}
> {color:#0747a6} } catch (final DataStoreException e) {{color}
> {color:#0747a6} log.error("S3 Connection could not be
> created." + e);{color}
> {color:#0747a6} throw new DataStoreException("S3 Connection
> could not be created");{color}
> {color:#0747a6} }{color}
> {color:#0747a6} }{color}
> {color:#172b4d}Even after setting collectLimit in code, still it is taking
> 100000 as limit.{color}
> {color:#172b4d}Expected :{color}
> versionGCOptions.collectLimit should set to custom value to avoid
> DocumentStoreException or solution to avoid DocumentStoreException if
> documents exceeded to 100000.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
