gogrlx/nats-server
1
0
Fork 0
mirror of https://github.com/gogrlx/nats-server.git synced 2026-04-02 11:48:43 -07:00
Code Issues Packages Projects Releases Wiki Activity
5,390 Commits 30 Branches 48 Tags
0cc7302be9bc37f0ee45f480e4e4a458b3bb62bf
Commit Graph

3530 Commits

Author SHA1 Message Date
Derek Collison
0cc7302be9 A stream name is tied to its identity and can not be changed on a restore. 
Signed-off-by: Derek Collison <derek@nats.io>
 2022-02-09 12:38:45 -08:00
Derek Collison
c13a84cf44 Fixed a bug that would calculate the first sequence of a filteredPending incorrectly. 
Also added in more optimized version to select the first matching message in a message block for LoadNextMsg.

Signed-off-by: Derek Collison <derek@nats.io>
 2022-02-08 13:29:38 -08:00
Derek Collison
d50febeeff Improved sparse consumers replay time. 
When a stream has multiple subjects and a consumer filters the stream to a small and spread out list of messages the logic would do a linear scan looking for the next message for the filtered consumer.
This CL allows the store layer to utilize the per subject info to improve the times.

Signed-off-by: Derek Collison <derek@nats.io>
 2022-02-07 17:26:32 -08:00
Derek Collison
d71bb63e1b Merge pull request #2847 from nats-io/fc_stall 
Fixed flow control stall under specific conditions of message size.
 2022-02-07 09:28:17 -08:00
Derek Collison
55b7f11c9a Fixed flow control stall under specific conditions of message size. 
Signed-off-by: Derek Collison <derek@nats.io>
 2022-02-05 20:15:48 -08:00
Ivan Kozlovic
6d92051696 Release v2.7.2 
Signed-off-by: Ivan Kozlovic <ivan@synadia.com>
 2022-02-04 14:49:39 -07:00
Derek Collison
a0a2e32185 Remove dynamic account behaviors. 
We used these in tests and for experimenting with sandboxed environments like the demo network.

Signed-off-by: Derek Collison <derek@nats.io>
 2022-02-04 13:32:18 -08:00
Derek Collison
5da0453964 Add in NumSubjects to StreamInfo 
Signed-off-by: Derek Collison <derek@nats.io>
 2022-02-02 08:51:13 -08:00
Derek Collison
6a3cf0f71e Added in ability to get number of subjects from StreamInfo, and optionally details per subject on how many messages each subject has. 
This can also be filtered, meaning you can filter out the subjects when asking for details.

Signed-off-by: Derek Collison <derek@nats.io>
 2022-02-02 08:51:13 -08:00
Derek Collison
12f5ea3655 When a consumer had not filtered subject and was attached to a interest policy retention stream we could incorrectly drop messages. 
Signed-off-by: Derek Collison <derek@nats.io>
 2022-02-01 14:21:05 -08:00
Derek Collison
6b5332249b This test was using fetch and failing if the complete batch was not filled. 
This has nothing to do with the test, we just want to make sure the leader steps down and there were no low level errors on the fetch.

Signed-off-by: Derek Collison <derek@nats.io>
 2022-02-01 13:34:00 -08:00
Derek Collison
8fce45dfa7 Under certain scenarios the pending for a consumer could appear to get stuck. 
Under the covers we were calculating pending per msg block incorrectly when a single message existed beyond the requested sequence.

Signed-off-by: Derek Collison <derek@nats.io>
 2022-02-01 12:17:08 -08:00
Ivan Kozlovic
30c431a9a3 [FIXED] JetStream: BackOff redeliveries would always use first in list 
If the consumer's sequence was not the same than the stream's sequence,
then the redelivery would always use the first duration from the
BackOff list.

Signed-off-by: Ivan Kozlovic <ivan@synadia.com>
 2022-01-31 17:44:08 -07:00
Derek Collison
0d158728d1 Merge pull request #2824 from nats-io/fix-nodeinfo 
Store JetStream Config in node info map
 2022-01-31 13:58:12 -08:00
Derek Collison
fa814f7cee Fixed behavior for when MaxMsgsPerSubject is set and DiscardNew is also set. 
Signed-off-by: Derek Collison <derek@nats.io>
 2022-01-31 08:36:37 -08:00
Derek Collison
6b690bd5b6 Bump to 2.7.2-beta 
Signed-off-by: Derek Collison <derek@nats.io>
 2022-01-30 15:24:07 -08:00
Derek Collison
8815072e34 Fix flapping test 
Signed-off-by: Derek Collison <derek@nats.io>
 2022-01-30 14:54:24 -08:00
Derek Collison
b38ced51b2 A true no wait pull request was not considering redeliveries. 
Signed-off-by: Derek Collison <derek@nats.io>
 2022-01-30 11:52:35 -08:00
Derek Collison
275d42628b Fix for #2828. The original design of the consumer and the subsequent store did not allow updates. 
Now that we do, we need to store the new config into our storage layer.

Signed-off-by: Derek Collison <derek@nats.io>
 2022-01-30 09:45:05 -08:00
Derek Collison
a57bd96def Updating a push consumer to be pull would succeed but cause a panic if used. 
This disallows that upgrade. We had a check in place for pull to push, but not the reverse.

Signed-off-by: Derek Collison <derek@nats.io>
 2022-01-28 13:11:58 -08:00
Jaime Piña
ae8eedb88e Store JetStream Config in node info map 2022-01-27 14:46:41 -08:00
Ivan Kozlovic
37ff12cc7a Release v2.7.1 
Signed-off-by: Ivan Kozlovic <ivan@synadia.com>
 2022-01-25 13:54:12 -07:00
Ivan Kozlovic
7f81f2d4c6 Merge pull request #2816 from nats-io/revocation-issue-442 
Fix jwt based user/activation token revocation and granularity
 2022-01-25 13:42:14 -07:00
Matthias Hanel
c5cc81bc1d Fix collectRevocations 
Signed-off-by: Matthias Hanel <mh@synadia.com>
 2022-01-25 15:02:47 -05:00
Matthias Hanel
274ec6db65 incorporate comments 
Signed-off-by: Matthias Hanel <mh@synadia.com>
 2022-01-25 14:53:03 -05:00
Matthias Hanel
fa12d096cd Fix jwt based user/activation token revocation and revocation granularity 
user and activation token did not honor the jwt value for all * on
connect.

activation token where not re evaluated when the export revoked a key.
In part this is a consistency measure so servers that already have an
account and servers that don't behave the same way.

in jwt activation token revocations are stored per export.
The server stored them per account, thus effectively merging
revocations. Now they are stored per export inside the server too.

fixes nats-io/nsc/issues/442

Signed-off-by: Matthias Hanel <mh@synadia.com>
 2022-01-25 13:48:12 -05:00
Derek Collison
6486cd8fc8 Added in /healthz endpoint for health and liveness probes in environments like k8s. 
Currently this code returns a 200 and { "status": "ok" } iff all configured ports are open
and if JetStream is configured and we have contact with the metaleader and the cluster and all streams are up to date.

Signed-off-by: Derek Collison <derek@nats.io>
 2022-01-24 19:30:10 -08:00
Derek Collison
6be9925127 Update config error 
Signed-off-by: Derek Collison <derek@nats.io>
 2022-01-24 15:02:41 -08:00
Derek Collison
65b168aa8b Updates based on feedback. MaxDeliver needs to be set properly now to be > len(BackOff) but if larger we will reuse last value in BackOff array. 
Signed-off-by: Derek Collison <derek@nats.io>
 2022-01-24 15:02:39 -08:00
Derek Collison
bd78b1a99b Formal json version for NAK delay 
Signed-off-by: Derek Collison <derek@nats.io>
 2022-01-24 15:01:52 -08:00
Derek Collison
d486c24199 Allow a consumer to be configured with BackOffs. 
This allows a consumer to have exponential backoffs vs static AckWait and MaxDeliver.
When BackOff is set it will overridde AckWait to BackOff[0] and MaxDeliver will be len(BackOff)+1.

Signed-off-by: Derek Collison <derek@nats.io>
 2022-01-24 14:57:36 -08:00
Derek Collison
579bf336ad Allow NAK to take a delay parameter to delay redelivery for a certain amount of time. 
Signed-off-by: Derek Collison <derek@nats.io>
 2022-01-24 14:57:28 -08:00
Derek Collison
89435d50b2 Merge pull request #2813 from nats-io/pull_consumer_2 
Updates to Pull Consumers
 2022-01-24 13:52:16 -08:00
Derek Collison
d332684322 Fixed data race and fuxed bug that we would not clear our waiting queue when a leader stepped down. 
Signed-off-by: Derek Collison <derek@nats.io>
 2022-01-24 13:01:25 -08:00
Derek Collison
35e8c57f06 Merge pull request #2814 from nats-io/neg_deleted 
If we detect negative deleted, adjust to 0.
 2022-01-24 11:52:58 -08:00
Ivan Kozlovic
dc3972304d Merge pull request #2791 from mattstep/mattstep/replication-lag-metrics 
[ADDED] Replication lag to stream monitoring
 2022-01-24 12:07:06 -07:00
Derek Collison
d07000cde0 If we detect negative deleted, adjust 
Signed-off-by: Derek Collison <derek@nats.io>
 2022-01-24 10:52:46 -08:00
Derek Collison
6fd41e5ea4 Updates based on review feedback 
Signed-off-by: Derek Collison <derek@nats.io>
 2022-01-24 10:23:47 -08:00
Derek Collison
d962500827 Track reply subjects for pending pull requests across clustered consumers. 
We will only send if all peers in our group are >= 2.7.1 and we will check for updates.
When a consumer follower takes over it will notify all pending requests that those requests are invalid now.

Signed-off-by: Derek Collison <derek@nats.io>
 2022-01-21 16:31:59 -08:00
Matt Stephenson
59cc0f0015 Add source and mirror info to stream monitoring 2022-01-21 12:44:42 -08:00
Derek Collison
ecd0d8eaff Reset authSet state 
Signed-off-by: Derek Collison <derek@nats.io>
 2022-01-21 11:03:19 -08:00
Derek Collison
ae43b24be4 Fixed #2810. Do not warn about password for internally generated no auth user when just system account assigned to non-default. 
Signed-off-by: Derek Collison <derek@nats.io>
 2022-01-21 10:09:10 -08:00
Derek Collison
dcadf6643a When no_auth_user is set, do not require a CONNECT proto. 
Signed-off-by: Derek Collison <derek@nats.io>
 2022-01-21 09:41:45 -08:00
Derek Collison
7f572983ac The 2.7 update broke the one-shot pull consumer fetch behavior due to change and a fix to a bug that allowed it to work before. 
This change tries to lock down all expected behaviors, and now does out of order timeouts for requests.

Signed-off-by: Derek Collison <derek@nats.io>
 2022-01-20 18:06:29 -08:00
Ivan Kozlovic
7aba8a8e9e Bump version to v2.7.1-beta02 
Signed-off-by: Ivan Kozlovic <ivan@synadia.com>
 2022-01-20 16:32:30 -07:00
Jaime Piña
c82b583d7a Fix race condition in HTTP monitoring shutdown (#2805) 2022-01-20 15:29:17 -08:00
Ivan Kozlovic
f6acc9d28b [FIXED] Possible deadlock 
This is due to a re-entrant RLock(). It works sometimes, but if there
is a go routine requesting the write lock, then the second RLock()
will not be granted which will lead to a deadlock.
In summary: one should never make re-entrant RLock calls.

Signed-off-by: Ivan Kozlovic <ivan@synadia.com>
 2022-01-20 13:53:52 -07:00
Ivan Kozlovic
c76bf81267 Fixed goreleaser ldflags 
Since we moved to go modules, the path must include the `/v2`.

Signed-off-by: Ivan Kozlovic <ivan@synadia.com>
 2022-01-13 19:02:41 -07:00
Ivan Kozlovic
f1d2f7c797 Release v2.7.0 
Signed-off-by: Ivan Kozlovic <ivan@synadia.com>
 2022-01-13 17:11:02 -07:00
Ivan Kozlovic
70b1f5c950 Fix goreleaser script invocation - bump to rc2 
Signed-off-by: Ivan Kozlovic <ivan@synadia.com>
 2022-01-13 16:58:19 -07:00