New Features and Improvements in VictoriaMetrics - Q3 2025
VictoriaMetrics
0 views
29 slides
Oct 13, 2025
Slide 1 of 29
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
About This Presentation
New 7 releases in Q3
Search in documentation - now with AI
vmui - new alerting page
Dashboards: alert statistics -available at https://grafana.com/orgs/victoriametrics/dashboards
Alerting best practices - read at https://victoriametrics.com/blog/alerting-best-practices
Metrics usage tracker enabled...
New 7 releases in Q3
Search in documentation - now with AI
vmui - new alerting page
Dashboards: alert statistics -available at https://grafana.com/orgs/victoriametrics/dashboards
Alerting best practices - read at https://victoriametrics.com/blog/alerting-best-practices
Metrics usage tracker enabled by default, v1.122.0
Log new series toggle, v1.123.0
Usability improvements:
* share queries for debugging
* export raw data
* labels limit per target
Performance optimizations
VictoriaTraces - database for storing traces, see https://victoriametrics.com/blog/dev-note-distributed-tracing-with-victorialogs/
Opentelemetry-demo: logs, metrics and traces with VM - see more https://github.com/VictoriaMetrics-Community/opentelemetry-demo
What's new in VictoriaMetrics
Roman Khavronenko
Q3 2025
LTS:
●v1.110 - v1.110.19
●v1.122 - v1.122.4 - new LTS line
●v1.102 - deprecated
New 7 releases in Q3
Upstream:
●v1.120.0 - v1.126.0
Search in documentation
Search in documentation
Search in documentation - now with AI! ✨??????????????????????????????
Try it now at https://docs.victoriametrics.com/
Search in documentation
Thanks to @makasim, @andrewchubatiuk, @tenmozes and @kacperwalter!
vmui - new alerting page, v1.125.0
vmui - new alerting page
vmui - new alerting page
Thanks to @andrewchubatiuk, @haleygo, @arturminchukov and @loori-r!
Try it now at https://play.victoriametrics.com/
Dashboards: alert statistics
Available at https://grafana.com/orgs/victoriametrics/dashboards
Dashboards: alert statistics
Try it now at https://play-grafana.victoriametrics.com/
Alerting best practices
Read at https://victoriametrics.com/blog/alerting-best-practices
Metrics usage tracker enabled by default, v1.122.0
Log new series toggle, v1.123.0
1.Call /internal/log_new_series?seconds=60 on vmstorage or vmsingle
2.Check the logs:
2025-10-02T08:40:38 info enabling logging of new series for the next 1m0s. This may increase resource usage during this period.
2025-10-02T08:40:39 info new series created:
vminsert_request_duration_seconds_bucket{job="victoria-metrics",instance="self",vmrange="1.136e-06...1.292e-06"}
2025-10-02T08:40:39 info new series created: vminsert_request_duration_seconds_sum{job="victoria-metrics",instance="self"}
2025-10-02T08:40:39 info new series created: vminsert_request_duration_seconds_count{job="victoria-metrics",instance="self"}
…
3. Logging disables automatically in `?seconds=<duration>`
Log new series toggle, v1.123.0
Thanks to @leiwingqueen!
Usability improvements
vmsingle and vmstorage in VictoriaMetrics cluster : The -snapshotsMaxAge flag default
have been changed to 3d.
This enables automatic deletion of snapshots older than 3 days.
If you want to keep the previous behavior (never automatically deleting snapshots), please set
-snapshotsMaxAge=0.
See #9344 for the details.
Usability improvements: share queries for debugging
Usability improvements: export raw data
Usability improvements: export raw data
●Export in Json or CSV format
●Raw data is essential for troubleshooting
●https://docs.victoriametrics.com/victoriametrics/troubl
eshooting/#unexpected-query-results
●Raw data can be easily imported into VM
Usability improvements: labels limit per target
vmagent and Single-node VictoriaMetrics : support label_limit scrape parameter.
It sets labels limit for each scraped sample from a target. If limit is exceeded, the scrape will be
rejected and target marked with corresponding scrape error.
The label_limit can be defined on global level, per job or during relabeling via
__label_limit__ label.
See #7660 and #3233 issues.
Performance optimizations
FEATURE: improve performance on systems with many CPU cores by removing the top sources of false sharing at global variables. See
#8682 . Thanks to @tIGO for raising this issue and for the initial attempt to fix it at the PR #8683 .
FEATURE: vmselect in VictoriaMetrics cluster : dynamically adjusts the concurrent dial limit between 8 and 64 based on
-search.maxConcurrentRequests . Additionally, goroutines now have the opportunity to access available connections while
awaiting the dial limit token. This enables faster connection establishment when sudden requests arrive and reduces the blocking time
during the availability check for connections. See #8922
FEATURE: vmsingle and vmstorage in VictoriaMetrics cluster : reduce CPU usage caused by garbage collection during indexDB
lookups on cache misses. See #9324 for details.
FEATURE: vmagent and vmsingle : improve Kubernetes service discovery performance with large amount of configured role selectors.
See this #9354 issue for details. Thanks to the @fxrlv
FEATURE: vmsingle and vmstorage in VictoriaMetrics cluster : optimize /api/v1/labels and /api/v1/label/TAG/values
requests with a single match or extra_filters filter containing time series metric name. See this PR #9489 for details.
One more thing…
One more thing…
VictoriaTraces - database for storing traces
●Accepts trace spans in OpenTelemetry protocol
●Supports Jaeger Query Service JSON APIs:
○Integrates with Grafana via Jaeger datasource
○Integrates with Jaeger UI
●Uses up to 3.7x less RAM and up to 2.6x less CPU than other solutions such as
Grafana Tempo
●Supports alerting via vmalert
VictoriaTraces - database for storing traces
Developer's Note: Research on Distributed Tracing,
Comparing With Tempo and ClickHouse
VictoriaTraces - database for storing traces
VictoriaTraces - database for storing traces
Opentelemetry-demo: logs, metrics and traces with VM