-
Notifications
You must be signed in to change notification settings - Fork 22
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
Add metrics to monitor the query costs #51
base: main
Are you sure you want to change the base?
Conversation
- Add keepAlive flag to not kill the app when the query fail
@guerremdq -- adding these metrics would be welcomed. Thank you! Please be mindful of unit tests, code coverage, and the build status. |
Hey @stephen-soltesz , Need to implement a few more methods in the cloudtest library to be able to test this |
Changes from m-lab/go#174 are available in https://github.com/m-lab/go/releases/tag/v0.1.67 |
* - Add metrics for sucessful and failed queries - Add keepAlive flag to not kill the app when the query fail * - Add metric names prefix - Remove duplicated matric calls - Add extra label to update duration metic
Thanks @stephen-soltesz , I let me know what you think about this PR |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thank you @guerremdq -- I've added a few suggestions that I think will make the change a little simpler.
Reviewed 1 of 7 files at r1, 1 of 7 files at r3, 1 of 1 files at r5, 2 of 2 files at r7.
Reviewable status: 0 of 1 LGTMs obtained
query/bigquery_runner.go
line 22 at r7 (raw file):
} func (b *bigQueryImpl) Query(query string, visit func(row map[string]bigquery.Value) error) (*bigquery.QueryStatistics, error) {
Rather than returning bigquery.QueryStatistics
here (and for all existing cases calling Query), consider extending the BQRunner interface instead.
Then the bigQueryImpl
struct could include a bigquery.QueryStatistics
field that's set after a successful query and could be read by callers that need it from a new method, say, LastStats()
or similar.
Then the *Collector.Update()
loop would call Query()
followed by LastStats()
to update the metrics -- and all the return type changes throughout this PR are unnecessary.
query/bigquery_runner.go
line 31 at r7 (raw file):
} if job == nil {
It should not be possible for Query.Run to return a nil err and a nil job. Please remove this check.
query/bigquery_runner.go
line 38 at r7 (raw file):
status, err := job.Wait(context.Background()) if err != nil { return nil, status.Err()
status
is not guaranteed to be non-nil when err != nil
. Prefer simply returning err
here.
query/bigquery_runner.go
line 44 at r7 (raw file):
it, err := job.Read(context.Background())
nit: please remove extra new line between job.Read and the err check.
sql/collector.go
line 155 at r7 (raw file):
func setSlotMillis(ch chan<- prometheus.Metric, slotMillis int64, metricName string) { desc := prometheus.NewDesc("bqx_slot_seconds_utilized", "slot milliseconds utilized", []string{"filename"}, nil)
If these metrics were part of the Collector they would need to be registered in Describe, iirc.
Each set of Collector metrics may be distinct from one another. These metrics are constant for the package. So, I don't think these should be part of the Collector. These are package level metrics collected about the different queries run.
Please redefine these at the package level as standard Gauges.
.travis.yml
line 14 at r6 (raw file):
- make - go test -short -v ./... -cover=1 -coverprofile=_c.cov - $GOPATH/bin/goveralls -service=travis-ci -coverprofile=_c.cov
If you'd be willing to change "travis-ci" to "travis-pro" we may be able to get coverage stats for this PR. (But it might not work until a new PR is created)
Expose metrics for each query to monitor the query cost
This change is