Author: chetanm
Date: Mon Mar 20 12:49:14 2017
New Revision: 1787755
URL: http://svn.apache.org/viewvc?rev=1787755&view=rev
Log:
OAK-5946 - Document indexing flow (WIP)
Add details around index lease
Modified:
jackrabbit/oak/trunk/oak-doc/src/site/markdown/query/indexing.md
Modified: jackrabbit/oak/trunk/oak-doc/src/site/markdown/query/indexing.md
URL:
http://svn.apache.org/viewvc/jackrabbit/oak/trunk/oak-doc/src/site/markdown/query/indexing.md?rev=1787755&r1=1787754&r2=1787755&view=diff
==============================================================================
--- jackrabbit/oak/trunk/oak-doc/src/site/markdown/query/indexing.md (original)
+++ jackrabbit/oak/trunk/oak-doc/src/site/markdown/query/indexing.md Mon Mar 20
12:49:14 2017
@@ -133,13 +133,36 @@ own bookkeeping of checkpoint state and
Prior to Oak 1.4 there was only one indexing lane `async`. In Oak 1.4 support
was added to create 2 lanes `async` and
`fulltext-async`. With 1.6 its possible to [create multiple
lanes](#async-index-setup).
+#### <a name="cluster"></a> Clustered Setup
+
+In a clustered setup it needs to be ensured by the host application that async
indexing jobs for specific lanes are to
+be run as singleton in the cluster. If `AsyncIndexUpdate` for same lane gets
executed concurrently on different cluster
+nodes then it can lead to race conditions where old checkpoint gets lost
leading to reindexing of the indexes.
+
+Refer to [clustering](../clustering.html#scheduled-jobs) for more details on
how the host application should schedule
+such indexing jobs
+
+##### <a name="async-index-lease"></a> Indexing Lease
+
+`AsyncIndexUpdate` has an inbuilt lease logic to ensure that even if the jobs
gets scheduled to run on different cluster
+nodes then also only one of them runs. This is done by keeping a lease
property which gets periodically updated as
+indexing progresses.
+
+An `AsyncIndexUpdate` run would skip indexing if current lease has not expired
i.e. if the last
+update of lease was done long ago (default 15 mins) then it would be assumed
that cluster node doing indexing is not
+available and some other node would take over.
+
+The lease logic can delay start of indexing if the system is not stopped
cleanly. As of Oak 1.6 this does not affect
+non clustered setup like those based on SegmentNodeStore but only [affects
DocumentNodeStore][OAK-5159] based setups
+
#### <a name="async-index-lag"></a> Indexing Lag
#### <a name="async-index-setup"></a> Setup
`Since 1.6`
-#### <a name="cluster"></a> Clustered Setup
+
+
#### <a name="async-index-mbean"></a> Clustered Setup
@@ -151,6 +174,7 @@ Prior to Oak 1.4 there was only one inde
### Lucene Indexes
+[OAK-5159]: https://issues.apache.org/jira/browse/OAK-5159
\ No newline at end of file
