Hello! During the CDS Reef session we discussed a few ideas related to code cleanup and complexity reduction. This e-mail is first from a longer series. Today I would like to ask about the `MemDB` implementation of our key-value database abstraction. **Does anybody use it?** There is a PR from 2020 [1] that squeezes the MemDB completely. For two years no user has been spotted while benefits from the removal are broader than just the kv subsystem. MemDB is the very last user of the bufferlist's cloning concept which imposes a run-time cost to any buffer instance in the system. Any objection for chopping ;-) the stuff off? Regards, Radek [1]: https://github.com/ceph/ceph/pull/36282 _______________________________________________ Dev mailing list -- dev@xxxxxxx To unsubscribe send an email to dev-leave@xxxxxxx
