History log of /3.0.3-GA/ep-engine/ (Results 1 - 25 of 4208)
Revision (<<< Hide revision tags) (Show revision tags >>>)Date Author Comments
(<<< Hide modified files)
(Show modified files >>>)
Revision tags: v4.6.5, v4.6.3, v4.6.4, v4.6.2, v4.6.1, v4.6.0, v4.6.0-DP, v4.5.1, v4.1.2, v3.1.6, v4.5.0, v4.1.1, v3.1.5, v3.1.4, v3.1.3, v4.1.0, v3.1.2, v3.1.1, v4.0.0, v3.1.0
f6f7782f13-Feb-2015 Mike Wiederhold <mikewied06@gmail.com>

[BP] MB-13359: Get DCP next snapshot if there are meta items, but no mutations

If we only get a snapshot end message when we call getItemsForCursor
then we will consider the snapshot to

[BP] MB-13359: Get DCP next snapshot if there are meta items, but no mutations

If we only get a snapshot end message when we call getItemsForCursor
then we will consider the snapshot to be empty and pause the stream.
Since a snapshot end message can only be in a closed checkpoint this
means that it's possible that we are pausing the stream for no
reason.

Change-Id: I29b8603287b41401fd6f5c1e4d4f185611d5b583
Reviewed-on: http://review.couchbase.org/47087
Reviewed-by: abhinav dangeti <abhinav@couchbase.com>
Tested-by: Michael Wiederhold <mike@couchbase.com>
Reviewed-on: http://review.couchbase.org/47962
Tested-by: abhinav dangeti <abhinav@couchbase.com>
Reviewed-by: Michael Wiederhold <mike@couchbase.com>

show more ...

6cd22a5a06-Mar-2015 abhinavdangeti <abhinav@couchbase.com>

MB-13757: Acquire snapshot lock before updating highSeq & snapshot seqs

Acquire snapshot lock before updating highSeqno and then the snapshot
sequence numbers for TAP, to avoid the flush

MB-13757: Acquire snapshot lock before updating highSeq & snapshot seqs

Acquire snapshot lock before updating highSeqno and then the snapshot
sequence numbers for TAP, to avoid the flusher racing with queueDirty
where only highSeqno is updated but not the snapshot sequence numbers.

Change-Id: I2d7c5932c1d4bda316047236218f4ca9336946a3
Reviewed-on: http://review.couchbase.org/47811
Reviewed-by: abhinav dangeti <abhinav@couchbase.com>
Tested-by: abhinav dangeti <abhinav@couchbase.com>

show more ...

f6af5d3f05-Mar-2015 abhinavdangeti <abhinav@couchbase.com>

MB-13792: Acquire lock before changing checkpointManager's lastBySeqno

Change-Id: I55e381418b0a5b89704f2e9912caabfa4df8d15c
Reviewed-on: http://review.couchbase.org/47665
Reviewed-by

MB-13792: Acquire lock before changing checkpointManager's lastBySeqno

Change-Id: I55e381418b0a5b89704f2e9912caabfa4df8d15c
Reviewed-on: http://review.couchbase.org/47665
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Reviewed-by: Michael Wiederhold <mike@couchbase.com>
Tested-by: abhinav dangeti <abhinav@couchbase.com>

show more ...

e75f5cb324-Feb-2015 abhinavdangeti <abhinav@couchbase.com>

Merge remote-tracking branch 'gerrit/3.0.1' into 3.0.3

+ MB-13386: Ensure that purging the highSeqno doesn't happen

Change-Id: Ieacb1d86d747a275181bb238145fd206a14be140


b4279b6024-Feb-2015 abhinavdangeti <abhinav@couchbase.com>

MB-13386: Ensure that purging the highSeqno doesn't happen

+ The highest seqno is to not be purged for DCP.
+ This change ensures that items are still queued
for deletion if found to

MB-13386: Ensure that purging the highSeqno doesn't happen

+ The highest seqno is to not be purged for DCP.
+ This change ensures that items are still queued
for deletion if found to be expired.

Change-Id: I8102d5f61989523efc4f39b70b225c05cdd1b128
Reviewed-on: http://review.couchbase.org/47241
Reviewed-by: Manu Dhundi <manu@couchbase.com>
Reviewed-by: Michael Wiederhold <mike@couchbase.com>
Tested-by: abhinav dangeti <abhinav@couchbase.com>

show more ...

a391b53920-Feb-2015 abhinavdangeti <abhinav@couchbase.com>

Fix build breakage on ubuntu

Change-Id: Ic8d7b9c17086a612b55e2c84a26259081d29c8d4
Reviewed-on: http://review.couchbase.org/47107
Reviewed-by: Manu Dhundi <manu@couchbase.com>
Tes

Fix build breakage on ubuntu

Change-Id: Ic8d7b9c17086a612b55e2c84a26259081d29c8d4
Reviewed-on: http://review.couchbase.org/47107
Reviewed-by: Manu Dhundi <manu@couchbase.com>
Tested-by: abhinav dangeti <abhinav@couchbase.com>

show more ...

a3f6175319-Feb-2015 Manu Dhundi <manu@couchbase.com>

MB-13479: Indicate rollback to DCP clients if purge_seqno > snap_start_seqno

The replica may not get all the items if there are purged items in the active
node. Hence this solution propo

MB-13479: Indicate rollback to DCP clients if purge_seqno > snap_start_seqno

The replica may not get all the items if there are purged items in the active
node. Hence this solution proposes to indicate a rollback to seqno 0 in case
the purge_seqno > snap_start_seqno.
Note: Since the occurrence of this scenario is pretty rare, rolling back to 0
may not be a bad thing.

Change-Id: I5c8403115110be136df5d4cb4e2704edc2a4c9e4
Reviewed-on: http://review.couchbase.org/47094
Tested-by: Manu Dhundi <manu@couchbase.com>
Reviewed-by: Michael Wiederhold <mike@couchbase.com>

show more ...

1cc8280d19-Feb-2015 Manu Dhundi <manu@couchbase.com>

Merge remote-tracking branch 'couchbase/3.0.1' into couchbase/3.0.3

Change-Id: I3d214e5b9b7a60831dcb685a6ef85a4ee0996f3e


f2bc069718-Feb-2015 Manu Dhundi <manu@couchbase.com>

MB-13386: Do not purge the item with highest sequence number from db.

When the highest seq number that DCP is supposed to read from
the db is the last seq number in the db and is purged,

MB-13386: Do not purge the item with highest sequence number from db.

When the highest seq number that DCP is supposed to read from
the db is the last seq number in the db and is purged,
DCP backfill does not know about it and hence waits for it.
This results in DCP connection hang.
To solve this problem, during compaction we do not purge the
item with last(highest) sequence number in the db.

Change-Id: Ib83c335da3f7c0a952e4b760309276f73bff4ccf
Reviewed-on: http://review.couchbase.org/47035
Reviewed-by: Michael Wiederhold <mike@couchbase.com>
Tested-by: Manu Dhundi <manu@couchbase.com>

show more ...

5fa8788016-Feb-2015 abhinavdangeti <abhinav@couchbase.com>

MB-10370: Replications' set/delWithMetas to use replication threshold

The setWithMetas and deleteWithMetas issue by consumers for
intra-cluster replication will need to work on tapThrott

MB-10370: Replications' set/delWithMetas to use replication threshold

The setWithMetas and deleteWithMetas issue by consumers for
intra-cluster replication will need to work on tapThrottleThreshold
as opposed to mutation_memory_threshold.

Change-Id: I576c9f9961e03e70430d58a192854c6faa14156d
Reviewed-on: http://review.couchbase.org/46907
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Tested-by: Chiyoung Seo <chiyoung@couchbase.com>

show more ...

e59abd3511-Feb-2015 abhinavdangeti <abhinav@couchbase.com>

MB-11527: Allow tuning of compaction_write_queue_cap

Configuration parameter: compaction_write_queue_cap
Can be set in runtime through cbepctl

Change-Id: Id35899865509f5de13a956

MB-11527: Allow tuning of compaction_write_queue_cap

Configuration parameter: compaction_write_queue_cap
Can be set in runtime through cbepctl

Change-Id: Id35899865509f5de13a9565802aeec7f84a71f3d
Reviewed-on: http://review.couchbase.org/46642
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Tested-by: abhinav dangeti <abhinav@couchbase.com>

show more ...

b398f0bf10-Feb-2015 abhinavdangeti <abhinav@couchbase.com>

MB-13205: Setting appropriate memory thresholds

mutation_mem_threshold: 93%
backfill_mem_threshold: 96%
replication_mem_threhold: 99%

Change-Id: Ic2abb80266c1e32d55e21e5359f

MB-13205: Setting appropriate memory thresholds

mutation_mem_threshold: 93%
backfill_mem_threshold: 96%
replication_mem_threhold: 99%

Change-Id: Ic2abb80266c1e32d55e21e5359fe0ce99e53551f
Reviewed-on: http://review.couchbase.org/46639
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Tested-by: abhinav dangeti <abhinav@couchbase.com>

show more ...

c1015f7110-Feb-2015 abhinavdangeti <abhinav@couchbase.com>

Conditionally delete expired items during compaction

Delete expired items during compaction if and only if
memory usage is lesser than threshold
(compaction_exp_mem_threshold) and di

Conditionally delete expired items during compaction

Delete expired items during compaction if and only if
memory usage is lesser than threshold
(compaction_exp_mem_threshold) and diskqueue size is
lesser than tap_throttle_queue_cap.

Change-Id: I256b127b32050dc0a1e395cacb369353a2fe0565
Reviewed-on: http://review.couchbase.org/46638
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Tested-by: abhinav dangeti <abhinav@couchbase.com>

show more ...

a3b0967f04-Feb-2015 Chiyoung Seo <chiyoung.seo@gmail.com>

[Backport] Disable the disk write queue size cap for replication backoff.

The disk write queue size cap for replication backoff was determined in
the very early versions of Couchbase Ser

[Backport] Disable the disk write queue size cap for replication backoff.

The disk write queue size cap for replication backoff was determined in
the very early versions of Couchbase Server that were deployed in
small-to-medium sized cluster with spining disks.

In our recent benchmark results, it turns out that disabling or setting
the disk write queue cap to a large value shows much better performance
behaviors in large-scale clusters with SSDs. As Couchbase Server is
deployed in such environments more and more, this change disables the
write queue size cap, but still supports a replication backoff by
checking the memory usage.

Note that the disk write queue size cap is still configurable at runtime.

Change-Id: Iedf711ad1c3bef61ca954f83f802b4a647b9ec88
Reviewed-on: http://review.couchbase.org/46367
Reviewed-by: Michael Wiederhold <mike@couchbase.com>
Tested-by: Chiyoung Seo <chiyoung@couchbase.com>
Reviewed-on: http://review.couchbase.org/46614
Reviewed-by: abhinav dangeti <abhinav@couchbase.com>

show more ...

4073e68004-Feb-2015 abhinavdangeti <abhinav@couchbase.com>

[Backport] MB-13286: New DcpProducer to have paused status set to true

DcpOpen will create a new DcpProducer and if one with a
similar name already exists (e.g - during rebalance),
t

[Backport] MB-13286: New DcpProducer to have paused status set to true

DcpOpen will create a new DcpProducer and if one with a
similar name already exists (e.g - during rebalance),
this new producer will replace the older one, and the
paused status will need to be set to true, to let the
notification for the new connection to be sent to
memcached.

Change-Id: I40c22601a7d29141741608339c58caa486a698b2
Reviewed-on: http://review.couchbase.org/46374
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Tested-by: abhinav dangeti <abhinav@couchbase.com>
Reviewed-on: http://review.couchbase.org/46511

show more ...

Revision tags: v3.0.2
fb165a6603-Dec-2014 abhinavdangeti <abhinav@couchbase.com>

MB-12673: Check for new checkpoint items after takeoverSend

Check for any new checkpoint items when the readyQ is found to
be empty in the takeoverSendPhase, before setting the old
v

MB-12673: Check for new checkpoint items after takeoverSend

Check for any new checkpoint items when the readyQ is found to
be empty in the takeoverSendPhase, before setting the old
vbucket to dead and the new vbucket to active.

Change-Id: I2a6ddacc711f5db42a1e3c575ae18d0b2b3126bd
Reviewed-on: http://review.couchbase.org/43918
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Tested-by: abhinav dangeti <abhinav@couchbase.com>

show more ...

5ad466ca24-Nov-2014 Trond Norbye <trond.norbye@gmail.com>

MB-12751: Return KEY EEXISTS for locked objects

Change-Id: I5eb8a24337a81fed861cc8140bbdcf1a132acd90
Reviewed-on: http://review.couchbase.org/43625
Reviewed-by: Chiyoung Seo <chiyoun

MB-12751: Return KEY EEXISTS for locked objects

Change-Id: I5eb8a24337a81fed861cc8140bbdcf1a132acd90
Reviewed-on: http://review.couchbase.org/43625
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Reviewed-by: Trond Norbye <trond.norbye@gmail.com>
Tested-by: Trond Norbye <trond.norbye@gmail.com>

show more ...

dfe4c89c14-Nov-2014 Sriram Ganesan <sriram@couchbase.com>

MB-12647: Ensure CAS value will always be unique

In windows, the gethrtime() returns the same timestamp when 2 requests
are made in a very short interval, thus causing the same CAS to re

MB-12647: Ensure CAS value will always be unique

In windows, the gethrtime() returns the same timestamp when 2 requests
are made in a very short interval, thus causing the same CAS to return.
This can race conditions with 2 consecutive requests, resulting in
data corruption.

Change-Id: I4f396de3f14129504ca406ebb8d4c7a9f3a89bd8
Reviewed-on: http://review.couchbase.org/43271
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Tested-by: Chiyoung Seo <chiyoung@couchbase.com>

show more ...

9ca901f811-Nov-2014 Sundar Sridharan <sundar.sridharan@gmail.com>

MB-12576: Limit writer threads to 4

Having more than 4 writers increases bgfetch latencies in DGM
This change selectively reverts 32a166c511d7b242433011a875402e1278300add
Change-Id:

MB-12576: Limit writer threads to 4

Having more than 4 writers increases bgfetch latencies in DGM
This change selectively reverts 32a166c511d7b242433011a875402e1278300add
Change-Id: Icdb996622237747e759c52751f2c8e613c9ba262
Reviewed-on: http://review.couchbase.org/43060
Tested-by: Sundararaman Sridharan <sundar@couchbase.com>
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>

show more ...

e8eab29406-Nov-2014 abhinavdangeti <abhinav@couchbase.com>

MB-12562: Avoid terminating all writer threads

With incremental writer thread scheduling, we
need to make sure that all writer threads aren't
deleted before deleting all buckets.

MB-12562: Avoid terminating all writer threads

With incremental writer thread scheduling, we
need to make sure that all writer threads aren't
deleted before deleting all buckets.

Scenario
- maxWriters is 4
- create bucket 1, numWriters=1
- create bucket 2, numWriters=2
- create bucket 3, numWriters=3
- create bucket 4, numWriters=4
- create bucket 5, numWriters=4
- delete bucket 5, numWriters=3 =>with change, 4
- delete bucket 4, numWriters=2 =>with change, 3
- delete bucket 3, numWriters=1 =>with change, 2
- delete bucket 2, numWriters=0 =>with change, 1
- delete bucket 1, numWriters=0 =>with change, 0

Change-Id: Ib9d45f7acb9177924612547538aa74ca9dd49c20
Reviewed-on: http://review.couchbase.org/42880
Reviewed-by: Sundararaman Sridharan <sundar@couchbase.com>
Tested-by: abhinav dangeti <abhinav@couchbase.com>

show more ...

3d867be106-Nov-2014 Mike Wiederhold <mike@couchbase.com>

MB-12551: Add all arguments to noop log message

Change-Id: Id6e8aac2bf9e398dc8c929eb122651884d27b7c1
Reviewed-on: http://review.couchbase.org/42862
Reviewed-by: Chiyoung Seo <chiyoun

MB-12551: Add all arguments to noop log message

Change-Id: Id6e8aac2bf9e398dc8c929eb122651884d27b7c1
Reviewed-on: http://review.couchbase.org/42862
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Tested-by: Michael Wiederhold <mike@couchbase.com>

show more ...

e187397503-Nov-2014 Mike Wiederhold <mike@couchbase.com>

MB-12483: Make sure vbucket_state initializes all variables

We need to make sure that this data is always completely filled
out so that we don't accidentally write garbage data to disk.

MB-12483: Make sure vbucket_state initializes all variables

We need to make sure that this data is always completely filled
out so that we don't accidentally write garbage data to disk.

Change-Id: I196e7ca9f5bada8e0df90ddb01b6e952650bed56
Reviewed-on: http://review.couchbase.org/42729
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Reviewed-by: abhinav dangeti <abhinav@couchbase.com>
Tested-by: Michael Wiederhold <mike@couchbase.com>

show more ...

8e87073b03-Nov-2014 Mike Wiederhold <mike@couchbase.com>

MB-12483: Remove the default constructor in the vbucket_state struct

The default constructor can leave uninitialized fields and we persist
this structure to disk. In oreder to prevent ga

MB-12483: Remove the default constructor in the vbucket_state struct

The default constructor can leave uninitialized fields and we persist
this structure to disk. In oreder to prevent garbage from being written
we should remove the default constructor.

Change-Id: I028fac1dc112bb454779a30f695eb180278455df
Reviewed-on: http://review.couchbase.org/42728
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Tested-by: Michael Wiederhold <mike@couchbase.com>

show more ...

7d037a7831-Oct-2014 Mike Wiederhold <mike@couchbase.com>

MB-12483: Don't allow recreation during vbucket deletion

When we delete a vbucket we have an option to recreate the file
immediately. Doing this is incorrect because we will not know

MB-12483: Don't allow recreation during vbucket deletion

When we delete a vbucket we have an option to recreate the file
immediately. Doing this is incorrect because we will not know
what the failover log of the new vbucket looks like until we
actually create it in memory. This can lead to a situation where
there is no failover log and as a result the local doc json will
be invalid. If the server is shutdown right after this happens and
then is restarted the vbuckets might be created with garbage values
in some of their fields.

Change-Id: I70e6335af68746aeac49a336da5e33b70dfcfe0e
Reviewed-on: http://review.couchbase.org/42677
Reviewed-by: Sundararaman Sridharan <sundar@couchbase.com>
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>
Tested-by: Michael Wiederhold <mike@couchbase.com>

show more ...

32a166c528-Oct-2014 Sundar Sridharan <sundar.sridharan@gmail.com>

MB-12279: Incrementally spawn writer threads for buckets

This change is needed to mitigate the high bgfetch latency
observed in heavy Data-Greater-than-Memory scenarios because
havin

MB-12279: Incrementally spawn writer threads for buckets

This change is needed to mitigate the high bgfetch latency
observed in heavy Data-Greater-than-Memory scenarios because
having high number of writing threads slows down disk read
performance.
Also mitigates MB-11143 slowdown in single HDD case

NOTE: This may slow down disk persistence in fast SSDs by default
NOTE: cbepctl can still be used to dynamically tune writers at runtime

Change-Id: Iddf0d3094f38b66ba8c0e09d6d6a307d15b38d56
Reviewed-on: http://review.couchbase.org/42552
Tested-by: Sundararaman Sridharan <sundar@couchbase.com>
Reviewed-by: Chiyoung Seo <chiyoung@couchbase.com>

show more ...

12345678910>>...169