History log of /3.0.3-GA/ep-engine/ (Results 326 - 350 of 4208)
Revision (<<< Hide revision tags) (Show revision tags >>>)Date Author Comments
(<<< Hide modified files)
(Show modified files >>>)
0aa9f23229-May-2014 Sundar Sridharan <sundar.sridharan@gmail.com>

MB-11232 Wake up tasks in pendingQueue as well

TaskQueue::wake() routine was missing to look at the
tasks blocked in the pendingQueue causing the
ExecutorPool::unregister() routine t

MB-11232 Wake up tasks in pendingQueue as well

TaskQueue::wake() routine was missing to look at the
tasks blocked in the pendingQueue causing the
ExecutorPool::unregister() routine to hang occasionally.

Change-Id: I3592a820ae7ef0d1c587a459f0f34ab6731bbdde
Reviewed-on: http://review.couchbase.org/37688
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Tested-by: Chiyoung Seo <chiyoung@couchbase.com>

show more ...

bdd4606629-May-2014 Sundar Sridharan <sundar.sridharan@gmail.com>

MB-11254 remove unnecessary assert unlockShardTask

This assert fires if unlockShardTask is called when the shard
is already unlocked.
If a shard-serialized writer task such as flushe

MB-11254 remove unnecessary assert unlockShardTask

This assert fires if unlockShardTask is called when the shard
is already unlocked.
If a shard-serialized writer task such as flusher gets
cancelled even before it got to run, it might end up
calling unlockShardTask() even though it never locked it.
This is ok - since the task is going away, so removing this assert.

Change-Id: Ife2c98611327e1d9d3e926b2252242231b2e48e5
Reviewed-on: http://review.couchbase.org/37689
Tested-by: Sundararaman Sridharan <sundar@couchbase.com>
Reviewed-by: Sriram Ganesan <sriram@couchbase.com>
Reviewed-by: Michael Wiederhold <mike@couchbase.com>

show more ...

8cc9f30328-May-2014 Sundar Sridharan <sundar.sridharan@gmail.com>

fix warning of unused endSeqno in BackfillDiskLoad

Change-Id: I678a38e3551d0260f5e7664675a3bd2fd0573726
Reviewed-on: http://review.couchbase.org/37661
Reviewed-by: Chiyoung Seo <chiy

fix warning of unused endSeqno in BackfillDiskLoad

Change-Id: I678a38e3551d0260f5e7664675a3bd2fd0573726
Reviewed-on: http://review.couchbase.org/37661
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Tested-by: Sundararaman Sridharan <sundar@couchbase.com>

show more ...

b697276329-May-2014 Sriram Ganesan <sriram@couchbase.com>

MB-11217:Use getShard() to retrieve shard for a vbucket id

When a rollback request is received for a particular vbucket id
on UprConsumer, we need to use getShard() function to get the

MB-11217:Use getShard() to retrieve shard for a vbucket id

When a rollback request is received for a particular vbucket id
on UprConsumer, we need to use getShard() function to get the
shard for a vbucket id

Change-Id: I6e69dcb50313e3d46a8021638b7fd25e6225279a
Reviewed-on: http://review.couchbase.org/37677
Tested-by: Sriram Ganesan <sriram@couchbase.com>
Reviewed-by: Michael Wiederhold <mike@couchbase.com>

show more ...

db2d005028-May-2014 Mike Wiederhold <mike@couchbase.com>

MB-11107: Ack at least every 5 secs to clear the upr flow ctl buffer

When there are no more ops being sent through replication we should
still ack whatever we have processed. If nothing

MB-11107: Ack at least every 5 secs to clear the upr flow ctl buffer

When there are no more ops being sent through replication we should
still ack whatever we have processed. If nothing has been acked
for 5 seconds we should send an ack.

Change-Id: I4248cd500d97e461797744fa682226ecde2c3a60
Reviewed-on: http://review.couchbase.org/37639
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Tested-by: Michael Wiederhold <mike@couchbase.com>

show more ...

7615577028-May-2014 Mike Wiederhold <mike@couchbase.com>

MB-11107: Add an acked_bytes stat to all upr consumers and upr producers

These stats will help with debugging upr flow control issues.

Change-Id: I58b93fc77ac4271d4072a7a27812740b50

MB-11107: Add an acked_bytes stat to all upr consumers and upr producers

These stats will help with debugging upr flow control issues.

Change-Id: I58b93fc77ac4271d4072a7a27812740b500bca3a
Reviewed-on: http://review.couchbase.org/37625
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Tested-by: Michael Wiederhold <mike@couchbase.com>

show more ...

2cd67ee028-May-2014 Mike Wiederhold <mike@couchbase.com>

MB-11085: Always create a new failover entry on unclean shutdowns

In the past we wouldn't generate a new failover entry if the high
seqno number on disk was the same after a crash. This

MB-11085: Always create a new failover entry on unclean shutdowns

In the past we wouldn't generate a new failover entry if the high
seqno number on disk was the same after a crash. This is incorrect
because it is possible that the server did receive mutations and
replicated them without persisting them before the crash. If this
happens the consumers of upr streams will no roll back their data
properly because the failover entry will not change on the server.

Change-Id: I8c6bab504f0be3298e1e888dbe6f3fac9c3fa905
Reviewed-on: http://review.couchbase.org/37670
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Tested-by: Michael Wiederhold <mike@couchbase.com>

show more ...

e0f06d6328-May-2014 Mike Wiederhold <mike@couchbase.com>

MB-11219: Always send a snapshot marker even if the snapshot has no items

It is possible that for a seqno range there are no items. Even if
this happens we still need to send a snapshot

MB-11219: Always send a snapshot marker even if the snapshot has no items

It is possible that for a seqno range there are no items. Even if
this happens we still need to send a snapshot marker to signify
what range the engine is trying to send. This is particularly a
problem when we do a vbucket takeover and there are items to send
during the takeover. If the last snapshot sent was a disk snapshot
then the new master vbucket can be left in backfilled state and
this will cause problems for persistence.

Change-Id: I6b5fade5195395a5f7ac0fe722336d44175e5884
Reviewed-on: http://review.couchbase.org/37669
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Tested-by: Michael Wiederhold <mike@couchbase.com>

show more ...

26d8ad6f27-May-2014 Mike Wiederhold <mike@couchbase.com>

MB-11107: Give upr flow control stats more descriptive names

When debugging a flow control issue I often found myself confused
as to what some of the stats really meant. Changing the nam

MB-11107: Give upr flow control stats more descriptive names

When debugging a flow control issue I often found myself confused
as to what some of the stats really meant. Changing the names to
be more descriptive helped me to more quickly debug issues.

Change-Id: I079b79bb6c92d7ec7b5c61c3a3aa5ad888b6d979
Reviewed-on: http://review.couchbase.org/37624
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Tested-by: Michael Wiederhold <mike@couchbase.com>

show more ...

f8ebe39627-May-2014 Mike Wiederhold <mike@couchbase.com>

MB-11107: The amount of processed bytes in an upr consumer should be atomic

We update this variable from different threads and as a reuslt need to make
it atomic to prevent concurrency i

MB-11107: The amount of processed bytes in an upr consumer should be atomic

We update this variable from different threads and as a reuslt need to make
it atomic to prevent concurrency issue. This fixes an issue where the
amount of acked bytes on the consumer did not match up with the producer.

Change-Id: I991ad51826be97404aaf4b63a9d616ffde62508a
Reviewed-on: http://review.couchbase.org/37623
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Tested-by: Michael Wiederhold <mike@couchbase.com>

show more ...

970af0b928-May-2014 Mike Wiederhold <mike@couchbase.com>

MB-11053: Don't create a new failover entry for upr set vb state

We were incorrectly creating a new failover entry when there was
a clean handoff through an upr takeover stream. Whenever

MB-11053: Don't create a new failover entry for upr set vb state

We were incorrectly creating a new failover entry when there was
a clean handoff through an upr takeover stream. Whenever there is
a clean handoff we should not generate a new failover entry.

Change-Id: If36055274e867032cc752407eff8aee39bff32ba
Reviewed-on: http://review.couchbase.org/37643
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Tested-by: Michael Wiederhold <mike@couchbase.com>

show more ...

db48e06728-May-2014 Chris Hillery <ceej@couchbase.com>

MB-11106: handle paths with spaces

Change-Id: Ide05143124ce41737d7b845af9c4ef22b006d0bc
Reviewed-on: http://review.couchbase.org/37631
Tested-by: Chris Hillery <ceej@couchbase.com>

MB-11106: handle paths with spaces

Change-Id: Ide05143124ce41737d7b845af9c4ef22b006d0bc
Reviewed-on: http://review.couchbase.org/37631
Tested-by: Chris Hillery <ceej@couchbase.com>
Reviewed-by: Trond Norbye <trond.norbye@gmail.com>

show more ...

8ca60c1321-May-2014 Sriram Ganesan <sriram@couchbase.com>

MB-11010: reduce lock contention in processBufferedItems()

In processBufferedItems(), responses are processed one at a time
from a particular stream's buffer. This leads to a lot of

MB-11010: reduce lock contention in processBufferedItems()

In processBufferedItems(), responses are processed one at a time
from a particular stream's buffer. This leads to a lot of
contention on the streamMutex protecting the stream. Reduce the
contention by batch-processing the responses. Currently the batch
size is arbitrarily set to 10.

Change-Id: If42e0e1aa208be381e116f78293ef749e1509143
Reviewed-on: http://review.couchbase.org/37568
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Tested-by: Chiyoung Seo <chiyoung@couchbase.com>

show more ...

fc21382e24-May-2014 Sundar Sridharan <sundar.sridharan@gmail.com>

Replace auxUnderlying with shard-level instance

Without this change we run the risk of race conditions
where threads that are supposed to work on their own shard
level CouchKVStore i

Replace auxUnderlying with shard-level instance

Without this change we run the risk of race conditions
where threads that are supposed to work on their own shard
level CouchKVStore instance end up working on the same
auxUnderlying instance.

Makes read operations like warmup about 10% faster!

Change-Id: Iaa171edc00eccfbbe92ac29c7b41f2e89316d6d4
Reviewed-on: http://review.couchbase.org/37549
Reviewed-by: David Liao <david.liao@couchbase.com>
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Tested-by: Chiyoung Seo <chiyoung@couchbase.com>

show more ...

a2b7e17326-May-2014 Trond Norbye <trond.norbye@gmail.com>

Keep track of number of item objects

Change-Id: I7a8a5d037f3f3b5243d1257cb400a8030d125a9d
Reviewed-on: http://review.couchbase.org/37583
Reviewed-by: Chiyoung Seo <chiyoung@couchbase

Keep track of number of item objects

Change-Id: I7a8a5d037f3f3b5243d1257cb400a8030d125a9d
Reviewed-on: http://review.couchbase.org/37583
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Tested-by: Chiyoung Seo <chiyoung@couchbase.com>

show more ...

1cb55ca022-May-2014 Trond Norbye <trond.norbye@gmail.com>

Add more statistics information for StoredValue objects

Track the number of them, their size and overhead

Change-Id: Ibafde84fd950c76ec9878fa903bf593836886b54
Reviewed-on: http:

Add more statistics information for StoredValue objects

Track the number of them, their size and overhead

Change-Id: Ibafde84fd950c76ec9878fa903bf593836886b54
Reviewed-on: http://review.couchbase.org/37433
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Tested-by: Chiyoung Seo <chiyoung@couchbase.com>

show more ...

5d93307727-May-2014 Trond Norbye <trond.norbye@gmail.com>

MB-11209: Document blob related stats

Change-Id: I3c262363202edc1a601b7d0f188f7fdde257bddf
Reviewed-on: http://review.couchbase.org/37604
Reviewed-by: Chiyoung Seo <chiyoung@couchbas

MB-11209: Document blob related stats

Change-Id: I3c262363202edc1a601b7d0f188f7fdde257bddf
Reviewed-on: http://review.couchbase.org/37604
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Tested-by: Chiyoung Seo <chiyoung@couchbase.com>

show more ...

8a277c1f22-May-2014 Trond Norbye <trond.norbye@gmail.com>

Add const for objectregistry

Change-Id: I686b1f4241d92b05b059a329f3c80241551610a1
Reviewed-on: http://review.couchbase.org/37434
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>

Add const for objectregistry

Change-Id: I686b1f4241d92b05b059a329f3c80241551610a1
Reviewed-on: http://review.couchbase.org/37434
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Tested-by: Trond Norbye <trond.norbye@gmail.com>

show more ...

71ff82fb22-May-2014 Trond Norbye <trond.norbye@gmail.com>

Use the actual allocated size for the blobs

Change-Id: Icdbb5267d831482a985ed628c1b2d97d29c45131
Reviewed-on: http://review.couchbase.org/37430
Reviewed-by: Chiyoung Seo <chiyoung@co

Use the actual allocated size for the blobs

Change-Id: Icdbb5267d831482a985ed628c1b2d97d29c45131
Reviewed-on: http://review.couchbase.org/37430
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Tested-by: Trond Norbye <trond.norbye@gmail.com>

show more ...

12260ea526-May-2014 Trond Norbye <trond.norbye@gmail.com>

Fixup check for unistd.h

Change-Id: I55411336d0ab81c2a9d33bcf854b400af9c857a9
Reviewed-on: http://review.couchbase.org/37582
Reviewed-by: Trond Norbye <trond.norbye@gmail.com>
Te

Fixup check for unistd.h

Change-Id: I55411336d0ab81c2a9d33bcf854b400af9c857a9
Reviewed-on: http://review.couchbase.org/37582
Reviewed-by: Trond Norbye <trond.norbye@gmail.com>
Tested-by: Trond Norbye <trond.norbye@gmail.com>

show more ...

74d9688d23-May-2014 David Liao <liaodw@gmail.com>

MB-11097 stats failovers should not have failovers in the result

Removed extra "failovers" string from stat failovers result.

Change-Id: I8c99b4c1d8ff97a3d779e6894aa3bbe629510ed3

MB-11097 stats failovers should not have failovers in the result

Removed extra "failovers" string from stat failovers result.

Change-Id: I8c99b4c1d8ff97a3d779e6894aa3bbe629510ed3
Reviewed-on: http://review.couchbase.org/37482
Reviewed-by: Sriram Ganesan <sriram@couchbase.com>
Reviewed-by: Michael Wiederhold <mike@couchbase.com>
Tested-by: Michael Wiederhold <mike@couchbase.com>

show more ...

7b07c48124-May-2014 Chiyoung Seo <chiyoung.seo@gmail.com>

Fix to the wrong start byseq num from checkpoint registration

Change-Id: I5a95025750d93600088b51d87f5294dbf2e9c644
Reviewed-on: http://review.couchbase.org/37567
Reviewed-by: Chiyoun

Fix to the wrong start byseq num from checkpoint registration

Change-Id: I5a95025750d93600088b51d87f5294dbf2e9c644
Reviewed-on: http://review.couchbase.org/37567
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Tested-by: Chiyoung Seo <chiyoung@couchbase.com>

show more ...

ba774c3124-May-2014 Mike Wiederhold <mike@couchbase.com>

Fix broken upr unit tests due to back takeover estimates

This "+ 1" was deleted by accident in a previous patch.

Change-Id: I50fab65f80abce60b3b8c0691526cf92a9043528
Reviewed-on

Fix broken upr unit tests due to back takeover estimates

This "+ 1" was deleted by accident in a previous patch.

Change-Id: I50fab65f80abce60b3b8c0691526cf92a9043528
Reviewed-on: http://review.couchbase.org/37566
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Tested-by: Chiyoung Seo <chiyoung@couchbase.com>

show more ...

1de10a5e24-May-2014 Sundar Sridharan <sundar.sridharan@gmail.com>

MB-11193: use right shard-level kvstore in warmup

General use of getAuxUnderlying() ends up causing all the warmup
threads to share the same CouchKVStore instance. This can result
in

MB-11193: use right shard-level kvstore in warmup

General use of getAuxUnderlying() ends up causing all the warmup
threads to share the same CouchKVStore instance. This can result
in corruption of shared datastructures like dbFileRevMap.
Use the appropriate shard level KVStore instance for separation.
Change-Id: Iccbfc44c50e6381890392f4b50f02d2d6c23ab28
Reviewed-on: http://review.couchbase.org/37510
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Tested-by: Chiyoung Seo <chiyoung@couchbase.com>

show more ...

a0d3117722-May-2014 Chiyoung Seo <chiyoung.seo@gmail.com>

MB-11104 Return the correct start seq number on cursor registration

When the cursor is registered, the start seq number to be returned
to the caller should be equal to the next item to b

MB-11104 Return the correct start seq number on cursor registration

When the cursor is registered, the start seq number to be returned
to the caller should be equal to the next item to be fetched
thorough a cursor.

Change-Id: I3f9aa8e40be6e5f3dafd3c218f17cb837bb4d96d
Reviewed-on: http://review.couchbase.org/37465
Reviewed-by: Michael Wiederhold <mike@couchbase.com>
Tested-by: Chiyoung Seo <chiyoung@couchbase.com>

show more ...

1...<<11121314151617181920>>...169