Please explain the data flow of sentry.

[hyunkj42]

There are more PODs and I am not sure what role they play.
Can you explain the sentry data flow (configuration diagram)?
Please briefly explain what operation is performed in which POD. :)

[pavels]

Well, there are multiple sources

Basically, https://github.com/getsentry/self-hosted and https://github.com/getsentry/self-hosted/blob/master/docker-compose.yml specifically is the first source of truth about what needs to run with what parameters. If it is not clear from there, you will probably need to dig through sentry source code.

There is 2 main parts - sentry itself, which is the web ui, batch job processor and it uses postgres as main datasource. The second part is Snuba which is part data processing part data storage and retrieval framework which uses clickhouse as data store. Check https://getsentry.github.io/snuba/architecture/overview.html for some details.

Snuba and sentry is than stitched together with kafka, celery (jobs in python) and redis in kind of a weird way. Some containers consuming or producing nto kafaka are sentry and some are snuba - see https://github.com/getsentry/self-hosted/blob/master/docker-compose.yml

If you go through Sentry commits and pull request you will find out, that they are kind of struggling with rate at which their containers multiply with each new feature. See for example getsentry/snuba#1670

[hyunkj42]

Thank you for answer.
I will take note. :)

[KamilKopaczyk]

If you need to understand the general architecture of Sentry, this is the page i often go back to: https://develop.sentry.dev/architecture

[Mokto]

This issue is stale because it has been open for 30 days with no activity.