History log of /6.0.3/ns_server/ (Results 26 - 50 of 13787)
Revision (<<< Hide revision tags) (Show revision tags >>>)Date Author Comments
(<<< Hide modified files)
(Show modified files >>>)
776e8ee604-Sep-2019 Aliaksey Artamonau <aliaksiej.artamonau@gmail.com>

MB-35748 Add a type spec to async_loop_with_result/1.

Otherwise dialyzer starts complaining about the next change.

In order for the return type to be no_return(), I also had to chan

MB-35748 Add a type spec to async_loop_with_result/1.

Otherwise dialyzer starts complaining about the next change.

In order for the return type to be no_return(), I also had to change
handle_get_result/2 to do exit(normal) after replying to the
caller. This was the effective behavior anyway, since
async_loop_with_result/1 is only called in a tail position in the
controlling process.

Change-Id: Ib8e196facb1c945e87a537cb1c8f8138eb496c12
Reviewed-on: http://review.couchbase.org/114288
Well-Formed: Build Bot <build@couchbase.com>
Tested-by: Aliaksey Artamonau <aliaksey.artamonau@couchbase.com>
Reviewed-by: Artem Stemkovski <artem@couchbase.com>

show more ...

bda056d604-Sep-2019 Aliaksey Artamonau <aliaksiej.artamonau@gmail.com>

MB-35748 Don't expect "future" requests.

Change-Id: Idf24c5e771897e7bb10090ac4446b6424823f26c
Reviewed-on: http://review.couchbase.org/114287
Well-Formed: Build Bot <build@couchbase.

MB-35748 Don't expect "future" requests.

Change-Id: Idf24c5e771897e7bb10090ac4446b6424823f26c
Reviewed-on: http://review.couchbase.org/114287
Well-Formed: Build Bot <build@couchbase.com>
Tested-by: Aliaksey Artamonau <aliaksey.artamonau@couchbase.com>
Reviewed-by: Artem Stemkovski <artem@couchbase.com>

show more ...

2a5a1fbd03-Sep-2019 Aliaksey Artamonau <aliaksiej.artamonau@gmail.com>

MB-35748 Drop unused monitor_pids async option.

Change-Id: Id9d1ec4c02fbc0bde6142f2ba6f50257c47875d1
Reviewed-on: http://review.couchbase.org/114286
Well-Formed: Build Bot <build@cou

MB-35748 Drop unused monitor_pids async option.

Change-Id: Id9d1ec4c02fbc0bde6142f2ba6f50257c47875d1
Reviewed-on: http://review.couchbase.org/114286
Well-Formed: Build Bot <build@couchbase.com>
Tested-by: Aliaksey Artamonau <aliaksey.artamonau@couchbase.com>
Reviewed-by: Artem Stemkovski <artem@couchbase.com>

show more ...

d1db8d4f03-Sep-2019 Aliaksey Artamonau <aliaksiej.artamonau@gmail.com>

MB-35748 Drop unused async:adopt/1.

Change-Id: Ia0ebcf2d10bd0ddaaa57c9c7300e5c8e67a7e728
Reviewed-on: http://review.couchbase.org/114285
Well-Formed: Build Bot <build@couchbase.com>

MB-35748 Drop unused async:adopt/1.

Change-Id: Ia0ebcf2d10bd0ddaaa57c9c7300e5c8e67a7e728
Reviewed-on: http://review.couchbase.org/114285
Well-Formed: Build Bot <build@couchbase.com>
Tested-by: Aliaksey Artamonau <aliaksey.artamonau@couchbase.com>
Reviewed-by: Artem Stemkovski <artem@couchbase.com>

show more ...

007a594831-Aug-2019 Aliaksey Artamonau <aliaksiej.artamonau@gmail.com>

MB-35748 Fix rebalance stop when bucket mover exits on its own.

If rebalance stop request is sent just at the moment when one of the
bucket movers terminates, the stop request may get ig

MB-35748 Fix rebalance stop when bucket mover exits on its own.

If rebalance stop request is sent just at the moment when one of the
bucket movers terminates, the stop request may get ignored. The issue
is in wait_for_mover function. When it receives a stop request from
the parent, it sends an exit to the mover and simply calls itself
recursively to wait till the mover exits. The problem is when the
mover terminates independently with reason 'normal'. In this case,
instead of exiting, wait_for_mover just returns 'ok', which continues
rebalance. The solution is to disentangle waiting for the mover
termination normally and when a stop request was received.

Change-Id: I59bb0bcfadf8ae436f9bf371f292384b4653422a
Reviewed-on: http://review.couchbase.org/114284
Well-Formed: Build Bot <build@couchbase.com>
Tested-by: Aliaksey Artamonau <aliaksey.artamonau@couchbase.com>
Reviewed-by: Artem Stemkovski <artem@couchbase.com>

show more ...

a4d3a72428-Aug-2019 Eben M Haber <eben@couchbase.com>

MB-35686 - Remove hard-coded upgrade instructions link.

Instead we should use the link that is helpfully provided by PhoneHome.

Change-Id: I6f66455975f1a1a8580716394e62425d1867eb89

MB-35686 - Remove hard-coded upgrade instructions link.

Instead we should use the link that is helpfully provided by PhoneHome.

Change-Id: I6f66455975f1a1a8580716394e62425d1867eb89
Reviewed-on: http://review.couchbase.org/113922
Well-Formed: Build Bot <build@couchbase.com>
Reviewed-by: Eben Haber <eben@couchbase.com>
Tested-by: Eben Haber <eben@couchbase.com>

show more ...

77de57ad27-Aug-2019 Ajit Yagaty <ajit.yagaty@couchbase.com>

MB-34855: Fixup the logout audit ID.

A new audit has been introduced in master branch which conflicts
with the audit ID 8255 assigned for logout. So assigning an ID of
8256 for logou

MB-34855: Fixup the logout audit ID.

A new audit has been introduced in master branch which conflicts
with the audit ID 8255 assigned for logout. So assigning an ID of
8256 for logout to avoid the conflict.

Change-Id: I97fe64f7dcfc6ed471178218a4e71a3020222a55
Reviewed-on: http://review.couchbase.org/113920
Tested-by: Ajit Yagaty <ajit.yagaty@couchbase.com>
Well-Formed: Build Bot <build@couchbase.com>
Reviewed-by: Ajit Yagaty <ajit.yagaty@couchbase.com>
Reviewed-by: Steve Watanabe <steve.watanabe@couchbase.com>

show more ...

108c78da21-Aug-2019 Ajit Yagaty <ajit.yagaty@couchbase.com>

MB-34855: Audit when user logs out of UI.

Change-Id: I413ccea7733d5170ae98da73c517ccc946e3ade4
Reviewed-on: http://review.couchbase.org/113651
Well-Formed: Build Bot <build@couchbase

MB-34855: Audit when user logs out of UI.

Change-Id: I413ccea7733d5170ae98da73c517ccc946e3ade4
Reviewed-on: http://review.couchbase.org/113651
Well-Formed: Build Bot <build@couchbase.com>
Tested-by: Ajit Yagaty <ajit.yagaty@couchbase.com>
Reviewed-by: Artem Stemkovski <artem@couchbase.com>

show more ...

f1d753a505-Jun-2019 Aliaksey Artamonau <aliaksiej.artamonau@gmail.com>

Merge remote-tracking branch 'couchbase/vulcan' into alice

* couchbase/vulcan:
MB-34140: Unbreak scripts/rebuild_replicas.sh

Change-Id: I19fd27647d135d1b139239d46b45be4c42397f

Merge remote-tracking branch 'couchbase/vulcan' into alice

* couchbase/vulcan:
MB-34140: Unbreak scripts/rebuild_replicas.sh

Change-Id: I19fd27647d135d1b139239d46b45be4c42397fdc

show more ...


Revision tags: 5.5.5
46682b4b10-May-2019 Aliaksey Artamonau <aliaksiej.artamonau@gmail.com>

MB-34140: Unbreak scripts/rebuild_replicas.sh

The change at http://review.couchbase.org/#/c/70225/ removed
ns_config_rep:syncrhonize_remote/1.

Change-Id: I494271ea1c0ae53f3f3bdd

MB-34140: Unbreak scripts/rebuild_replicas.sh

The change at http://review.couchbase.org/#/c/70225/ removed
ns_config_rep:syncrhonize_remote/1.

Change-Id: I494271ea1c0ae53f3f3bdd34acad54ad007e14a3
Reviewed-on: http://review.couchbase.org/108998
Tested-by: Aliaksey Artamonau <aliaksey.artamonau@couchbase.com>
Tested-by: Dave Finlay <dave.finlay@couchbase.com>
Reviewed-by: Dave Finlay <dave.finlay@couchbase.com>
Well-Formed: Build Bot <build@couchbase.com>

show more ...

dd3d5ab018-Apr-2019 Aliaksey Artamonau <aliaksiej.artamonau@gmail.com>

MB-33750 Try working around the bug upon taking over mastership.

The theory is that many customers will see the issue in question when
upgrading. The fact they didn't encounter it before

MB-33750 Try working around the bug upon taking over mastership.

The theory is that many customers will see the issue in question when
upgrading. The fact they didn't encounter it before may mean two
things:

- No master transitions happened in the observable past.
- There were master transitions, but they got lucky.

The first scenario is a problematic one. The leader_lease_acquirer
workers have probably accumulated enough stale PIDs to trigger the
problem once the new master takes over.

This patch is a best effort attempt to prevent the upgrade from
triggering the issue. When the fixed node takes over mastership, it
will attempt to kill the leader_lease_acquirer process on the old
master. This may or may not kill all of leader_lease_acquirer's worker
processes: those that don't die should get stuck in the shutdown. But
at least they are not going to be affecting the operation of the new
master anymore.

Change-Id: Ia812cce034a5708014f3e4c0415c24628e5cf423
Reviewed-on: http://review.couchbase.org/108003
Tested-by: Aliaksey Artamonau <aliaksey.artamonau@couchbase.com>
Reviewed-by: Artem Stemkovski <artem@couchbase.com>

show more ...

27fa0ece19-Apr-2019 Aliaksey Artamonau <aliaksiej.artamonau@gmail.com>

Allow waiting longer than 5 seconds for mb_master to surrender.

The issue was that we'd rpc:call mb_master:master_node() on an old
master. Yet master_node() internally uses a fixed defau

Allow waiting longer than 5 seconds for mb_master to surrender.

The issue was that we'd rpc:call mb_master:master_node() on an old
master. Yet master_node() internally uses a fixed default timeout of 5
seconds. So even if we used a larger rpc:call timeout, it would get
invalidated by master_node() internal timeout. The solution is to use
gen_fsm:sync_send_all_state_event instead of the rpc:call.

Change-Id: I7ac290dfef2a197d6ed5e8c710ff37f4c2fd395b
Reviewed-on: http://review.couchbase.org/108091
Tested-by: Aliaksey Artamonau <aliaksey.artamonau@couchbase.com>
Reviewed-by: Dave Finlay <dave.finlay@couchbase.com>
Reviewed-by: Artem Stemkovski <artem@couchbase.com>

show more ...

6fde9d9519-Apr-2019 Aliaksey Artamonau <aliaksiej.artamonau@gmail.com>

Move actual mb_master invalidation into a sub-function.

Change-Id: I0c5218ed17b8ce5ee77d1cf702b422dd4836009e
Reviewed-on: http://review.couchbase.org/108090
Tested-by: Aliaksey Artam

Move actual mb_master invalidation into a sub-function.

Change-Id: I0c5218ed17b8ce5ee77d1cf702b422dd4836009e
Reviewed-on: http://review.couchbase.org/108090
Tested-by: Aliaksey Artamonau <aliaksey.artamonau@couchbase.com>
Reviewed-by: Dave Finlay <dave.finlay@couchbase.com>
Reviewed-by: Artem Stemkovski <artem@couchbase.com>

show more ...

Revision tags: 5.5.4
64df4e2b18-Apr-2019 Aliaksey Artamonau <aliaksiej.artamonau@gmail.com>

Remove some legacy code from mb_master.

Change-Id: Id55eedd5377c192c0e7620d9a284c9b5eca74e8b
Reviewed-on: http://review.couchbase.org/108001
Tested-by: Aliaksey Artamonau <aliaksey.a

Remove some legacy code from mb_master.

Change-Id: Id55eedd5377c192c0e7620d9a284c9b5eca74e8b
Reviewed-on: http://review.couchbase.org/108001
Tested-by: Aliaksey Artamonau <aliaksey.artamonau@couchbase.com>
Reviewed-by: Dave Finlay <dave.finlay@couchbase.com>
Reviewed-by: Artem Stemkovski <artem@couchbase.com>

show more ...

98fb43f318-Apr-2019 Aliaksey Artamonau <aliaksiej.artamonau@gmail.com>

Adjust master advertised version to 6.0.2.

Change-Id: I91ba7a4d69fc8b9052e5ab3cab6018db3f2bfcb5
Reviewed-on: http://review.couchbase.org/108002
Tested-by: Aliaksey Artamonau <aliakse

Adjust master advertised version to 6.0.2.

Change-Id: I91ba7a4d69fc8b9052e5ab3cab6018db3f2bfcb5
Reviewed-on: http://review.couchbase.org/108002
Tested-by: Aliaksey Artamonau <aliaksey.artamonau@couchbase.com>
Reviewed-by: Dave Finlay <dave.finlay@couchbase.com>

show more ...

4b1dc6af16-Apr-2019 Aliaksey Artamonau <aliaksiej.artamonau@gmail.com>

Turn 'normal' into 'shutdown' in misc:unlink_terminate/2.

Similarly to misc:terminate_and_wait/2. Often we propagate exit reason
from one process to other related processes. If care is n

Turn 'normal' into 'shutdown' in misc:unlink_terminate/2.

Similarly to misc:terminate_and_wait/2. Often we propagate exit reason
from one process to other related processes. If care is not taken, we
might attempt to terminate some process with reason 'normal'. But that
won't work and the process will continue running.

Change-Id: I22db889bee8b987e065462b84f68b9e9d15665b9
Reviewed-on: http://review.couchbase.org/107863
Tested-by: Aliaksey Artamonau <aliaksey.artamonau@couchbase.com>
Reviewed-by: Dave Finlay <dave.finlay@couchbase.com>

show more ...

Revision tags: 6.0.1, 5.5.3, 6.0.0, 5.5.2, 5.1.2, 5.5.1, 5.5.0
001798eb11-Jun-2018 Aliaksey Artamonau <aliaksiej.artamonau@gmail.com>

[bp] Use shutdown_master_sup in mb_master:terminate.

To ensure that we notify on leader_events that we're not the master
anymore.

Original-Change-Id: I76ac9d41f704f9e770d20dc08c

[bp] Use shutdown_master_sup in mb_master:terminate.

To ensure that we notify on leader_events that we're not the master
anymore.

Original-Change-Id: I76ac9d41f704f9e770d20dc08ce5e978943e4190
Original-Reviewed-on: http://review.couchbase.org/95434
Original-Tested-by: Aliaksey Artamonau <aliaksey.artamonau@couchbase.com>
Original-Well-Formed: Build Bot <build@couchbase.com>
Original-Reviewed-by: Artem Stemkovski <artem@couchbase.com>

This is in preparation to addressing MB-33750.

Change-Id: I6e277459e9b93f7399329b51bad4f8fecdd83495
Reviewed-on: http://review.couchbase.org/107761
Tested-by: Aliaksey Artamonau <aliaksey.artamonau@couchbase.com>
Reviewed-by: Dave Finlay <dave.finlay@couchbase.com>

show more ...

d365644511-Jun-2018 Aliaksey Artamonau <aliaksiej.artamonau@gmail.com>

[bp] Remove the timeout logic from mb_master:shutdown_master_sup.

It's already the case that the supervisor should terminate in bounded
time. This is achieved through the use of appropri

[bp] Remove the timeout logic from mb_master:shutdown_master_sup.

It's already the case that the supervisor should terminate in bounded
time. This is achieved through the use of appropriate supervisor
shutdown policies. Having an extra timeout here just negates them. And
in fact, it's not safe to brutally kill supervisors. This may lead to
the supervisor children not being properly terminated.

Original-Change-Id: Ie8098fa00a8b86e1c110590bdfeed6de15627a08
Original-Reviewed-on: http://review.couchbase.org/95433
Original-Tested-by: Aliaksey Artamonau <aliaksey.artamonau@couchbase.com>
Original-Well-Formed: Build Bot <build@couchbase.com>
Original-Reviewed-by: Artem Stemkovski <artem@couchbase.com>

This is in preparation to addressing MB-33750.

Change-Id: I4bd026bf353a9ffdcac4f5b29a8e5209eba2a3b3
Reviewed-on: http://review.couchbase.org/107760
Tested-by: Aliaksey Artamonau <aliaksey.artamonau@couchbase.com>
Reviewed-by: Dave Finlay <dave.finlay@couchbase.com>

show more ...

81c7eae518-Apr-2019 Aliaksey Artamonau <aliaksiej.artamonau@gmail.com>

Merge remote-tracking branch 'couchbase/vulcan' into merge

* couchbase/vulcan:
MB-33853 Adjust master advertised version to 5.5.4.

Change-Id: I943378da3ff97433cb5d0f7494c5fe8c

Merge remote-tracking branch 'couchbase/vulcan' into merge

* couchbase/vulcan:
MB-33853 Adjust master advertised version to 5.5.4.

Change-Id: I943378da3ff97433cb5d0f7494c5fe8c5669c2a6

show more ...


42b0c12e18-Apr-2019 Aliaksey Artamonau <aliaksiej.artamonau@gmail.com>

MB-33853 Adjust master advertised version to 5.5.4.

- In case it's an upgrade from 5.5.x, take over mastership early and
hope that if the issue is hit, at least that happens early.

MB-33853 Adjust master advertised version to 5.5.4.

- In case it's an upgrade from 5.5.x, take over mastership early and
hope that if the issue is hit, at least that happens early.

- 6.0.2 will attempt to workaround the issue upon taking over
mastership. This change will let it distinguish between vulnerable
and immune flavors of 5.0.x.

Change-Id: Id3a3ddc3cea88bb3aaed3ffdb08c4f593d60ade8
Reviewed-on: http://review.couchbase.org/107999
Tested-by: Aliaksey Artamonau <aliaksey.artamonau@couchbase.com>
Well-Formed: Build Bot <build@couchbase.com>
Reviewed-by: Dave Finlay <dave.finlay@couchbase.com>

show more ...

42cbb72218-Apr-2019 Aliaksey Artamonau <aliaksiej.artamonau@gmail.com>

Merge remote-tracking branch 'couchbase/vulcan' into merge

* couchbase/vulcan:
MB-33853 Don't use async:perform for leader_lease_acquire_worker.

Change-Id: I13b73fe99b50d2e606

Merge remote-tracking branch 'couchbase/vulcan' into merge

* couchbase/vulcan:
MB-33853 Don't use async:perform for leader_lease_acquire_worker.

Change-Id: I13b73fe99b50d2e60665728e5faa7b3bc85674d4

show more ...


a537789f29-Jan-2019 Aliaksey Artamonau <aliaksiej.artamonau@gmail.com>

MB-33853 Don't use async:perform for leader_lease_acquire_worker.

Use a regular spawn_link instead. The parent process traps exits
anyway, so it just needs to handle those instead of the

MB-33853 Don't use async:perform for leader_lease_acquire_worker.

Use a regular spawn_link instead. The parent process traps exits
anyway, so it just needs to handle those instead of the monitor DOWN
messages.

The problem in the referred bug is due to async:perform/1 not cleaning
up PIDs of temporary processes that leader_lease_acquire_worker
spawns (via misc:executing_on_new_process/1). Over a long time each
worker will accumulate a significant number of PIDs of processes that
has long been dead. Once the VM spawns enough new processes (on the
order of 2²⁸ in our case, but the number can be significantly lower in
certain situations), it starts recycling old PIDs. So what happens is
that a newly spawned processes might take over some of the PIDs that
some worker still holds on to. When such worker is terminated (most
likely because another node became a master), at least two undesirable
situations may arise:

- The worker terminates an unrelated process that has taken over one
of its PIDs.

- The worker might get stuck. This is possible because internally
misc:terminate_and_wait/2 first sends an exit signals to all PIDs
it's given and only then starts monitoring the PIDs to wait till
they terminate. So if one of the PIDs is overtaken right in between
these two steps, the worker will end up waiting for a process to
terminate when that process never received the exit signal. Hence
the worker gets stuck.

It's the second case that was observed in couple of customer
escalations. There are two factors that make it quite likely:

- Over time a worker will accumulate a lot of PIDs, so the window
between two steps in misc:terminate_and_wait/2 above becomes quite
large: each of the steps has to traverse this long list.

- Incidentally, long worker shutdown time (which slows down the
shutdown of mb_master_sup) causes mb_master process to get blocked
for some time. This in turn results in node_monitor and related
processes to crash and restart because mb_master:master_mode()
times out. So it's these processes that start at the exactly wrong
moment and lock up the entire shutdown sequence.

Change-Id: I47dd29bf970c636669b5eea44f87ff281fc9c176
Reviewed-on: http://review.couchbase.org/107989
Well-Formed: Build Bot <build@couchbase.com>
Tested-by: Aliaksey Artamonau <aliaksey.artamonau@couchbase.com>
Reviewed-by: Dave Finlay <dave.finlay@couchbase.com>
Reviewed-by: Artem Stemkovski <artem@couchbase.com>

show more ...

9b1b8d5411-Apr-2019 Ajit Yagaty <ajit.yagaty@couchbase.com>

MB-33764: Allow users, with security admin role...

...assigned, to POST an HTTP request to the reloadCertificate endpoint.

Currently, the uploadClusterCA endpoint can be accessed by

MB-33764: Allow users, with security admin role...

...assigned, to POST an HTTP request to the reloadCertificate endpoint.

Currently, the uploadClusterCA endpoint can be accessed by a user who
has "security admin" role assigned. But the same user can't access the
reloadCertificate endpoint as this is allowed only for users with
"cluster admin" role assigned which is incorrect.

This patch allows only the users with "security admin" role assigned
to access uploadClusterCA and reloadCertificate endpoints.

Change-Id: I09399f87098424cc884d6abe8078ac18559f9253
Reviewed-on: http://review.couchbase.org/107685
Tested-by: Ajit Yagaty <ajit.yagaty@couchbase.com>
Well-Formed: Build Bot <build@couchbase.com>
Reviewed-by: Artem Stemkovski <artem@couchbase.com>

show more ...

a46e946312-Mar-2019 Artem Stemkovski <astemkov@gmail.com>

MB-33321 trim the resulting chain in ns_janitor:sanify_chain

...if it appears to be longer than the original chain

Change-Id: Ib46e1eb5152aa23d19e562d443d50a6c7aa2fcde
Reviewed-

MB-33321 trim the resulting chain in ns_janitor:sanify_chain

...if it appears to be longer than the original chain

Change-Id: Ib46e1eb5152aa23d19e562d443d50a6c7aa2fcde
Reviewed-on: http://review.couchbase.org/106089
Well-Formed: Build Bot <build@couchbase.com>
Tested-by: Artem Stemkovski <artem@couchbase.com>
Reviewed-by: Aliaksey Artamonau <aliaksey.artamonau@couchbase.com>

show more ...

45148db910-Jan-2019 Artem Stemkovski <astemkov@gmail.com>

MB-31354: allow to configure new security headers, not only

...overwrite the default ones

Change-Id: I11ac841bcc604ef4783b3c3851d54620f90cc144
Reviewed-on: http://review.couchba

MB-31354: allow to configure new security headers, not only

...overwrite the default ones

Change-Id: I11ac841bcc604ef4783b3c3851d54620f90cc144
Reviewed-on: http://review.couchbase.org/103470
Reviewed-by: Ajit Yagaty <ajit.yagaty@couchbase.com>
Well-Formed: Build Bot <build@couchbase.com>
Tested-by: Ajit Yagaty <ajit.yagaty@couchbase.com>
Reviewed-on: http://review.couchbase.org/106153
Tested-by: Artem Stemkovski <artem@couchbase.com>
Reviewed-by: Artem Stemkovski <artem@couchbase.com>

show more ...

12345678910>>...552