This repository has been archived by the owner on Apr 2, 2024. It is now read-only.
Requirements
- PostgreSQL: 12.x, 13.x, 14.x., 15.x
- TimescaleDB: >= 2.9.1 single-node (multi-node is not supported in this release)
- Promscale Connector: 0.17.0
Changed
- Maintenance jobs were separated from one another. The behavior of the old
prom_api.config_maintenance_jobs
call changed and it sets identical parameters
for each job type (e.g. parallelism 2 means each job type has 2 jobs, not 2 jobs in total).
There is also the newprom_api.config_maintenance_jobs
that allows configuring
them individually. Finally, the new default worker allocation is 2 metrics retention
jobs, 3 jobs for metrics compression and 1 job for traces retention. [#555] [#588] - Removed disk size-related columns from
prom_info.metric
.
Addedprom_info.metric_detail
which includes the disk size-related columns.
Improved the performance of these. [#547] - The metric compression job now only locks metrics that are known to have chunks
that need to be compressed [#576] - Renamed
_ps_catalog.chunks_to_freeze
to_ps_catalog.compressed_chunks_to_freeze
[#595] - All docker images upgraded to use TimescaleDB 2.9 [#598]
- The extension is updated to add support for PostgreSQL 15. [#590]
Added
- Telemetry for active series and last updated [#534]
- Added
_ps_catalog.compressed_chunks_missing_stats
view [#595]
Fixed
- Column conflict when creating a metric view with a label called
series
[#559]