[SENTRY-1083] Improve sentry cache (server, client) by using a stable commercial or main-strain cache

• Type: New Feature
• Status: Open
• Priority: Major
• Resolution: Unresolved
• Affects Version/s: 1.7.0
• Component/s: Sentry
• Labels: supportability
• Assignee: Unassigned
• Reporter: Anne Yu
• Created: 19/Feb/16 06:21
• Updated: 19/Feb/16 06:58

Description

We've seen many unreproducible issues because the sentry caches (either server side or client/plugin side) are full, or out of sync; We could start to consider a main-stream or open source cache to use across sentry project. The benefit of a widely used cache system:

1. has basic monitoring or dump facility;
2. has useful logging system;
3. usually is very high performanced in the memory;

I know for example, kafka incorporate RocksDb (code example).

Issue Links

• relates to: SENTRY-1080 Uber Jira for improving/redesigning sentry generic model