This reverts commit 93887a9b82a71613067ff0706248c18fe131c5b0.
See discussion at https://github.com/prometheus/prometheus/pull/17240 - I had
thought this was a simple and clear bugfix, but multiple people were depending
on the previous behavior already, and we want to err on the side of stability.
Signed-off-by: Julius Volz <julius.volz@gmail.com>
The modifiers were already printed as part of the VectorSelector, so for:
`foo[5m] anchored`
...the output was:
`foo anchored[5m] anchored`
Similar to how it was already done for `@` and `offset`, I now removed these
modifiers in the copy of the vector selector that is used to print the matrix
selector. I also removed some unused code that restored the copy of the vector
selector after overwriting its fields. AFAICS there was no use in doing that,
since it was a copy already that would just be thrown away after printing, and
the original selector wasn't affected. I also removed an erroneous comment in
`atOffset()` where no actual copying took place and no fields were overwritten.
Signed-off-by: Julius Volz <julius.volz@gmail.com>
Fixes#17370
In Prometheus v3.7.0, using labelmap actions with replacement patterns
containing regex variables (e.g., `$1`, `${1}`) would fail validation
when `metric_name_validation_scheme` was set to `legacy`, causing
Prometheus to fail at startup with:
"$1" is invalid 'replacement' for labelmap action
This was a regression as the same configuration worked in v3.6.0.
The issue was in the validation logic: while UTF-8 validation correctly
allowed `$` characters, legacy validation incorrectly used
`IsValidLabelName` which rejects `$` characters. The fix ensures legacy
validation uses `relabelTargetLegacy` regex which explicitly supports
regex template variables.
Added test cases to verify labelmap validation works with both `$1` and
`${1}` replacement patterns under legacy validation scheme.
Signed-off-by: Julien Pivotto <291750+roidelapluie@users.noreply.github.com>
Add ANCHORED and SMOOTHED keywords to the maybe_label and
metric_identifier rules in the parser grammar, allowing them
to be used as metric names and label names, similar to other
keywords like 'offset', 'step', and 'bool'.
This fixes an issue where expressions like `anchored{job="test"}`
and `sum by (smoothed) (some_metric)` would fail to parse.
Signed-off-by: Julien Pivotto <291750+roidelapluie@users.noreply.github.com>
As a follow-up to #17344, add two configuration parameters for controlling label
name translation, both defaulting to on for backwards compatibility (currently
these behaviours are hardcoded as enabled):
* otlp.label_name_underscore_sanitization => Prefix label names starting with a
single underscore with key_ when translating OTel attribute names
* otlp.label_name_preserve_multiple_underscores => Keep multiple consecutive
underscores in label names when translating OTel attribute names
Signed-off-by: Arve Knudsen <arve.knudsen@gmail.com>
Add the "@" string representation for the AT token in ItemTypeStr map to
ensure proper token-to-string conversion.
Signed-off-by: Julien Pivotto <291750+roidelapluie@users.noreply.github.com>
Loading the local region from the Instance MetaData Service broke in v3.7. This adds the IMDS call back in order to load the local region when no other method has set the region.
fixes#17375
Signed-off-by: Joe Adams <github@joeadams.io>
Ensure params are joined with & and remove trailing dot to restore working links from Alerts/Rules to Graph.
Signed-off-by: Akshat Sinha <akshatsinhasramhardy@gmail.com>
Fixes#17255.
The implementation happens mostly in the Add and Sub method, but the reconciliation works for all relevant operations. For example, you can now `rate` over a range wherein the custom bucket boundaries are changing.
Any custom bucket reconciliation is flagged with an info-level annotation.
---------
Signed-off-by: Linas Medziunas <linas.medziunas@gmail.com>
Signed-off-by: Linas Medžiūnas <linasm@users.noreply.github.com>
Without `staleTime: Infinity`, the query data would be immediately marked
stale, and in combination with the `gcTime: 0` setting and a brief unmount
during data loading (via Suspense), the data would have to be reloaded a second
time during/after the first real render.
`gcTime: 0` + `staleTime: Infinity` should give us the desired behavior where
data loaded for a page should be valid forever as long as that page is being
displayed, but always thrown away (and later reloaded) whenever we navigate
away and again to that page.
Signed-off-by: Julius Volz <julius.volz@gmail.com>
After the upgrade to AWS SDK v2, the EC2 and Lightsail service discovery
stopped working when using the default AWS credential chain (environment
variables, IAM roles, EC2 instance metadata, etc.).
The issue was that the code unconditionally created a StaticCredentialsProvider
with empty credentials when access_key and secret_key were not configured. In
AWS SDK v2, this causes a "static credentials are empty" error and prevents
the SDK from falling back to its default credential chain.
Signed-off-by: Julien Pivotto <291750+roidelapluie@users.noreply.github.com>
The modifiers were already printed as part of the VectorSelector, so for:
`foo[5m] anchored`
...the output was:
`foo anchored[5m] anchored`
Similar to how it was already done for `@` and `offset`, I now removed these
modifiers in the copy of the vector selector that is used to print the matrix
selector. I also removed some unused code that restored the copy of the vector
selector after overwriting its fields. AFAICS there was no use in doing that,
since it was a copy already that would just be thrown away after printing, and
the original selector wasn't affected. I also removed an erroneous comment in
`atOffset()` where no actual copying took place and no fields were overwritten.
Signed-off-by: Julius Volz <julius.volz@gmail.com>
Add ANCHORED and SMOOTHED keywords to the maybe_label and
metric_identifier rules in the parser grammar, allowing them
to be used as metric names and label names, similar to other
keywords like 'offset', 'step', and 'bool'.
This fixes an issue where expressions like `anchored{job="test"}`
and `sum by (smoothed) (some_metric)` would fail to parse.
Signed-off-by: Julien Pivotto <291750+roidelapluie@users.noreply.github.com>
This callback is not used by Prometheus, but in downstream projects it
is wasteful to allocate an ID only to abandon it.
Remove lengthy commment which I feel is distracting from the flow.
Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
Refactor the code so that everything proceeds linearly.
Also renamed `getOrSet` to `setUnlessAlreadySet` to emphasise that the
caller is expecting it not to be set.
Signed-off-by: Bryan Boreham <bjboreham@gmail.com>
