* Contribute grafana/agent sigv4 code
* address review feedback
- move validation logic for RemoteWrite into unmarshal
- copy configuration fields from ec2 SD config
- remove enabled field, use pointer for enabling sigv4
* Update config/config.go
* Don't provide credentials if secret key / access key left blank
* Add SigV4 headers to the list of unchangeable headers.
* sigv4: don't include all headers in signature
* only test for equality in the authorization header, not the signed date
* address review feedback
1. s/httpClientConfigEnabled/httpClientConfigAuthEnabled
2. bearer_token tuples to "authorization"
3. Un-export NewSigV4RoundTripper
* add x-amz-content-sha256 to list of unchangeable headers
* Document sigv4 configuration
* add suggestion for using default AWS SDK credentials
Signed-off-by: Robert Fratto <robertfratto@gmail.com>
Co-authored-by: Julien Pivotto <roidelapluie@gmail.com>
* Fix inconsistent display of word 'ago' in last evaluation column on rules page
Signed-off-by: Alex Petrov <alex.petrov.vt@gmail.com>
* Extract adding word 'ago' to relativeDuration util function
Signed-off-by: Alex Petrov <alex.petrov.vt@gmail.com>
Currently hashmod hashes all the labels, even if there is a newline. To
ensure it still works in the future, let's add a test.
Signed-off-by: Julien Pivotto <roidelapluie@inuits.eu>
This PR introduces support for follow_redirect, to enable users to
disable following HTTP redirects.
Signed-off-by: Julien Pivotto <roidelapluie@inuits.eu>
* Move remote read handler to remote package.
This follows the pattern I started with the remote write handler. The api/v1 package is getting pretty cluttered. Moving code to other packages helps reduce this size and also makes it reusable - eg Cortex doesn't do streaming remote writes yet, and will very soon.
Signed-off-by: Tom Wilkie <tom.wilkie@gmail.com>
* Deal with a nil remoteReadHandler for tests.
Signed-off-by: Tom Wilkie <tom.wilkie@gmail.com>
* Remove the global metrics.
Signed-off-by: Tom Wilkie <tom.wilkie@gmail.com>
* Fix test.
Signed-off-by: Tom Wilkie <tom.wilkie@gmail.com>
* Review feedback.
Signed-off-by: Tom Wilkie <tom.wilkie@gmail.com>
* Add expression explorer
Signed-off-by: Lucas Hild <git@lucas-hild.de>
* Add final new line to all files
Signed-off-by: Lucas Hild <git@lucas-hild.de>
* Rename expression to metric
Signed-off-by: Lucas Hild <git@lucas-hild.de>
* Pass dedicated metrics array to metrics explorer
Signed-off-by: Lucas Hild <git@lucas-hild.de>
* Fix styling of button
Signed-off-by: Lucas Hild <git@lucas-hild.de>
* Use append instead of prepend
Signed-off-by: Lucas Hild <git@lucas-hild.de>
* Update max width of modal
Signed-off-by: Lucas Hild <git@lucas-hild.de>
* Fix code style
Signed-off-by: Lucas Hild <git@lucas-hild.de>
* Fix inconsistent variable naming
Signed-off-by: Lucas Hild <git@lucas-hild.de>
* Fix modal title
Signed-off-by: Lucas Hild <git@lucas-hild.de>
* Fix tests
Signed-off-by: Lucas Hild <git@lucas-hild.de>
* Prevent request from being cached
Signed-off-by: Lucas Hild <git@lucas-hild.de>
* Remove timestamp from request
Signed-off-by: Lucas Hild <git@lucas-hild.de>
* Update button selector in test
Signed-off-by: Lucas Hild <git@lucas-hild.de>
* Refactor passing down metric names and query history
Signed-off-by: Lucas Hild <git@lucas-hild.de>
* Fix code style
Signed-off-by: Lucas Hild <git@lucas-hild.de>
The main branch tests are not passing due to the fact that #8489 was not
rebased on top of #8007.
Signed-off-by: Julien Pivotto <roidelapluie@inuits.eu>
This moves the label lookup into TSDB, whilst still keeping the cached-ref optimisation for repeated Appends.
This makes the API easier to consume and implement. In particular this change is motivated by the scrape-time-aggregation work, which I don't think is possible to implement without it as it needs access to label values.
Signed-off-by: Tom Wilkie <tom.wilkie@gmail.com>
Accepting alert_rule_test without alertname is confusing as it will
always pass with empty exp_alerts, and never with non-empty exp_alerts.
Signed-off-by: Raphael Bauduin <raphael.bauduin@tessares.net>
Right now a new segment might be created unnecessarily if the
uncompressed record would not fit, but after compression (typically
reducing record size in half) it would.
Signed-off-by: Chris Marchbanks <csmarchbanks@gmail.com>
Julien Pivotto