-
-
Notifications
You must be signed in to change notification settings - Fork 61
[gocd] split snuba deploys into snuba-deploy-py and snuba-deploy-rs #7507
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
Conversation
volokluev
left a comment
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.
great start, I do think the migrations need to be removed from the rust stage
| --container-name="transactions-consumer-new" \ | ||
| --container-name="transactions-subscriptions-executor" \ | ||
| --container-name="transactions-subscriptions-scheduler" \ | ||
| --container-name="snuba" \ |
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.
what is this container?
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.
this is a dumb catch-all that just matches everything. If we keep it then it will restart all rust processes
| --container-name="cleanup" \ | ||
| --container-name="optimize" \ | ||
| --container-name="cardinality-report" |
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.
what's the cleanup job? Seems like you're also removing the cardinality report, which makes sense just unexpected in 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.
I'll add those back in. I was reading some GoCD output and thought they had been deleted but it looks like they may still be defined:
INFO:scripts.k8s.k8s_deploy:Using context gke_internal-sentry_us-central1-b_zdpwkxst.
INFO:scripts.k8s.k8s_deploy:🔎 Looking for cronjobs in namespace 'default', matching label selector 'service=snuba' and with container names ['cleanup', 'optimize', 'cardinality-report']
INFO:scripts.k8s.k8s_deploy:👍 Found 5 matching cronjob(s)
volokluev
left a comment
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.
approved but please make the group attributes change
| elastic_profile_id: 'snuba', | ||
| tasks: [ | ||
| gocdtasks.script(importstr '../bash/check-github.sh'), | ||
| gocdtasks.script(importstr '../bash/check-migrations.sh'), |
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.
don't think this stage is necessary
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.
I think we do want migration checks, because I think we want to block a consumer deploy on any migrations having not run in that SHA. If the python pipeline is blocked for some reason, a later rust commit could reach the "front of the line" for a single tenant.
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.
ah fair enough, I recommend putting that in a comment, it's not an intuitive conclusion
|
PR reverted: 5fcd22c |
…loy-rs (#7507)" This reverts commit fc66311. Co-authored-by: onewland <[email protected]>
…take 2) (#7510) Retrying #7507 which had to be reverted. Hopefully getsentry/devinfra-deployment-service#800 fixed the underlying configuration issue
Split out the current deploy pipelines into
deploy-snuba-pyanddeploy-snuba-rs. The goal of this is to set us up for a future where we can only deploy rust consumers when changes are made in the Rust path (reducing the amount of SLO breaches on high-traffic consumers for rebalances).At this stage, we'll still deploy both with any code change to test. Next, we'll want to change the triggering logic to ignore code changes from irrelevant sources