Pgcalc borked - https://github.com/ceph/ceph.io/issues/265 - https://lists.ceph.io/hyperkitty/list/ceph-users@xxxxxxx/thread/KA5L6MZZ7K7XLCSZPNU5NQULZOXOAV2U/#55UCVU67Y7SKJBKSRKRGXZ7GAMT6V42C https://old.ceph.com/pgcalc still works, update docs to point there. The pg autoscaler should make this unnecessary. libcephsqlite segfault: https://tracker.ceph.com/issues/57152 - being investigated Project Aquarium has been archived Feedback of registration taking too long to be approved to report bugs on the tracker - redmine upgrade - new VM ready and DB imported - may have captcha support so account approval could be automatic again - current queue is empty, and monitored in different timezones - may be able to setup login via github boost/valgrind followup: - further discussion on dev list under 'boost and valgrind' - waiting on leads to approve https://github.com/ceph/ceph-build/pull/2043 - now merged https://tracker.ceph.com/issues/57147 (pacific, OSD crash) - to be investigated compiling cephadm - details - https://pad.ceph.com/p/cephadm-refactoring - https://github.com/ceph/ceph/pull/41855 - cephadm binary is massive, nearly 10000 lines, this would allow us to split up the file to make it more maintainable - already passing teuthology tests last time it was run I believe - idea is to have cephadm be "built" along with other things - the "build" is just one script that essentially using zipapp to combine the python files into one blob as it is now - makes minimum python version for cephadm binary 3.5 if you want to build it yourself, but adds no other dependencies - mostly just want to bring attention to this as it will have an effect on how the binary is obtained in main and R release onward. - Need to make sure as part of the release process the built binary is put somewhere users can access it _______________________________________________ Dev mailing list -- dev@xxxxxxx To unsubscribe send an email to dev-leave@xxxxxxx
