We have entries up to 1.8MB, which compresses down to ~450kb.
This is backwards compatible because we compress new entries, and fall
back to uncompressed if we fail to decompress. That's the same approach
as what we already use for pendingdata.
Details
Details
- Reviewers
machniak - Group Reviewers
Syncroton Developers - Commits
- rS51d7965cad8b: Compress the data in syncroton_relations_state
Diff Detail
Diff Detail
- Repository
- rS syncroton
- Lint
Lint Not Applicable - Unit
Tests Not Applicable
Event Timeline
Comment Actions
I feel uneasy about using binary data in sql queries, without special blob handling, but I guess it should work until we're proven otherwise.
Comment Actions
I'm assuming it's ok because it seems to me it's the same approach as for pendingdata (same column type etc).