-
Notifications
You must be signed in to change notification settings - Fork 1.1k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Exemplar support for Cluster #6135
base: cluster
Are you sure you want to change the base?
Conversation
Signed-off-by: Ted Possible <ted_possible@cable.comcast.com>
Added Exemplars to marshaling/unmarshaling processes Signed-off-by: Ted Possible <ted_possible@cable.comcast.com>
Added Exemplars to addRowToTimeSeries (scrapework) Signed-off-by: Ted Possible <ted_possible@cable.comcast.com>
- didn't work when there was no tags - didn't work when there was whitespace between metric name and tag start Added OpenMetrics Header for Exemplars to the client Improved tests for Exemplars Added Exemplars to PendingSeries tests and modified sizes to accomdoate exemplars In scrapework improved to only add exemplars if they exist. Signed-off-by: Ted Possible <ted_possible@cable.comcast.com>
Signed-off-by: Ted Possible <ted_possible@cable.comcast.com>
Signed-off-by: Ted Possible <ted_possible@cable.comcast.com>
…of allocs Fixed performance issues with row.unmarshal Signed-off-by: Ted Possible <ted_possible@cable.comcast.com>
Refactored and Optimized tag parsing utilzing tagsPool storage Bugfixes for white space at the end of exemplar Added passing test case for missing spaces between # and { for exemplar Signed-off-by: Ted Possible <ted_possible@cable.comcast.com>
Signed-off-by: Ted Possible <ted_possible@cable.comcast.com>
Numerous format fixes - comments Optimized exemplar label copying in scrapework.addRowToTimeSeres Signed-off-by: Ted Possible <ted_possible@cable.comcast.com>
Refactored Openmetrics header to only be included if scrapeExemplars flag is set More comment formatting fixes Signed-off-by: Ted Possible <ted_possible@cable.comcast.com>
…uplicate exemplars remote writing. TODO: need better testing for this Signed-off-by: Ted Possible <ted_possible@cable.comcast.com>
…meseries Various formatting fixes Signed-off-by: Ted Possible <ted_possible@cable.comcast.com>
Signed-off-by: Ted Possible <ted_possible@cable.comcast.com>
…ved unnecessary Tags member from the struct since we can use tagsPool Signed-off-by: Ted Possible <ted_possible@cable.comcast.com>
…emplars on subsequent sends. We only need exemplars once per timeseries. Signed-off-by: Ted Possible <ted_possible@cable.comcast.com>
Signed-off-by: Ted Possible <ted_possible@cable.comcast.com>
Misc comment formatting Signed-off-by: Ted Possible <ted_possible@cable.comcast.com>
Signed-off-by: Ted Possible <ted_possible@cable.comcast.com>
TODO: Store exemplars in IndexDB Monthly Partition
…ol on WriteRequest - This must be a separate pool from labelsPool because protobuf fields can be out of order
Thanks for PR, but I think, there is no need at porting changes to the cluster branch. We'll doing it on our own later, those changes also ll be ported to the enterprise version by VM team. |
This PR is for support of OpenMetrics Exemplars all the way through the to the storage level. Since cluster and single-node are vastly different in their implementations of vmstorage I have chosen the one that best fits our use case which is the cluster version. There is more to come in this PR with IndexDB and mergeset as well as vmselect and updates to vmstorage API. |
This adds support for OpenMetrics exemplars implementation into vmcluster.