Skip to content

Release v2.9.22
Compare
Choose a tag to compare
@NATS-CI NATS-CI released this 06 Sep 20:31
· 2693 commits to main since this release
v2.9.22
528ee14
This commit was created on GitHub.com and signed with GitHub’s verified signature. The key has expired.
GPG key ID: 4AEE18F83AFDEB23
Expired
Learn about vigilant mode.

Changelog

Go Version

  • 1.20.8 (updated out-of-cycle since Go 1.19 is now EOL)

Dependencies

  • github.com/nats-io/jwt/v2 v2.5.0
  • golang.org/x/crypto v0.12.0
  • golang.org/x/sys v0.11.0

Improved

Monitoring

  • CORS Allow-Origin passthrough for monitoring server (#4423) Thanks to @mdawar for the contribution!

JetStream

  • Improve consumer scaling reliability with filters and cluster restart (#4404)
  • Send event on lame duck mode (LDM) to avoid placing assets on shutting down nodes (#4405)
  • Skip filestore tombstones if downgrade from 2.10 occurs (#4452)
  • Adjust delivered and waiting count when consumer message delivery fails (#4472)

Fixed

Config

  • Allow empty configs and fix JSON compatibility (#4394, #4418)
  • Remove TLS OCSP debug log on reload (#4453)

Monitoring

  • Fix Content-Type header when /healthz is not 200 OK (#4437) Thanks to @mdawar for the contribution!
  • Fix server /connz idle time sorting (#4463) Thanks to @mdawar for the contribution!
  • Interface conversion bug which could cause a panic when calling /ipqueuesz endpoint (#4477)

Leafnode

  • Fix race condition which could affect propagating interest over leafnode connections (#4464)

JetStream

  • Fix possible deadlock in checking for drift in the usage reporting when storing a message (#4411)
  • Durable pull consumers could get cleaned up incorrectly on leader change (#4412)
  • Moving an R1 stream could sometimes lose all messages (#4413)
  • Prevent peer-remove of an R1 stream which could result in the stream becoming orphaned (#4420)
  • Ensure consumer ack pending is less than max ack pending on state restore (#4427)
  • Ensure to reset election timer when catching up (#4428) Thanks to @yuzhou-nj for the report!
  • Auto step-down Raft leader if an entry is missing on a catchup request (#4432)
  • Fix PurgeEx with keep having deletes in blocks (#4431)
  • Update global subject index when message blocks expire (#4439)
  • Ensure max messages per subject is respected after update (#4446) Thanks to @anthonyjacques20 for the report!
  • Ignore and remove empty message blocks on rebuild (#4447)
  • Fix possible accounting discrepancy on message write (#4455)
  • Fix potential message duplication from stream sources when downgrading from 2.10 (#4454)
  • Check for checksum violations for all records before sequence processing (#4465)
  • Fix message block accounting (#4473)

Complete Changes

v2.9.21...v2.9.22

Contributors

mdawar, anthonyjacques20, and yuzhou-nj
Assets 47
Loading