Release 0.0.35
Enhancements
- Support removing server from a cluster using their unique ID using
nats server raft peer-remove
, also show server IDs in several reports - Support requesting
healthz
from servers usingnats server req healthz
- Support the new Discard New Per Subject policy for streams
- Show the server core count in
nats server list
- Makes the request payload available as
.Request
in templates used innats server reply
- Show the bucket size in
nats kv info
- Support filtering streams in
stream report
andstream ls
based on subject filters - Adds a new account level
nats account report statistics
command showing connection information at an account level nats sub
can now take a filter subject even when a stream is selected using--stream
- Default to the
all
start policy when adding consumers - Support performing account data purges using
nats server purge
- Adds a new
nats stream state
command that shows only the state, also--state
flag tonats stream info
- Support enabling Mirror Direct mode when adding a stream
- Support creating mirrors of Key-Value buckets to facilitate read replicas using
nats kv add --mirror
. The intended use case is to facilitate read replicas in a different domain such as when using a branch level Leaf Node - When using the
prometheus
output format innats server check
the namespace for created metrics is now configurable using--namespace
- The
nats bench
command can act against an existing bucket using the new--bucketname
flag - Various UX improvements
Removals
As announced in the previous release a few features are deprecated and removed
- Stream template support were removed
- Remove
nats governor
command - Remove
nats backup|restore
, usenats stream backup|restore
ornats account backup|restore
Bug Fixes
- Improve error handling in connections report
- Do not offer stream retention as an editable property
- Fix some progress bars
- Improve help for duration flags
- Avoid array out of bounds panics in the
nats traffic
command - Improve help output for the reply command
- Improve paged connection reports that would fail when using
nats account report connections
on an account with more than 1000 connections - When waiting for a stream to elect a new leader with the
nats stream cluster down
command we waited 10 Milliseconds instead of 5 seconds
New Contributors
- @rickardgranberg made their first contribution in #594
- @kozlovic made their first contribution in #602
Full Changelog: v0.0.34...v0.0.35