[DISCUSS] Allow namespace admins to clone snapshots created by them
Hi Everyone, Creating a snapshot requires table admin permissions. But cloning it requires global admin permissions unless the user owns the snapshot and wants to recreate the original table the snapshot was based on using the same table name. This puts unnecessary load on the few users having global admin permissions on the cluster. I would like to relax this rule a bit and allow the owner of the snapshot to clone it into any namespace where they have admin permissions regardless of the table name used. Please let me know what you think about this proposal. And if you find it acceptable which branch do you think this could land on. Thanks, Szabolcs Bukros
Re: [DISCUSS] Allow namespace admins to clone snapshots created by them
FYI the PR is https://github.com/apache/hbase/pull/4885 and https://issues.apache.org/jira/browse/HBASE-27493. the proposal seems to be, should we allow cloning snapshot to any namespace if they're not the global admin. logically, it should be fine because they're the admin for the namespace, and should be able to do whatever within that namespace. Thanks, Stephen On Mon, Nov 21, 2022 at 11:38 AM Szabolcs Bukros wrote: > > Hi Everyone, > > Creating a snapshot requires table admin permissions. But cloning it > requires global admin permissions unless the user owns the snapshot and > wants to recreate the original table the snapshot was based on using the > same table name. This puts unnecessary load on the few users having global > admin permissions on the cluster. I would like to relax this rule a bit and > allow the owner of the snapshot to clone it into any namespace where they > have admin permissions regardless of the table name used. > > Please let me know what you think about this proposal. And if you find it > acceptable which branch do you think this could land on. > > Thanks, > Szabolcs Bukros
Re: [DISCUSS] Quickly release 2.5.2 with hadoop3 binaries
There is still a critical problem, filed HBASE-27501. 张铎(Duo Zhang) 于2022年11月20日周日 23:00写道: > > After fixing the jackson CVEs, I think it is time to make a new release now. > > Will put up 2.5.2RC0 next week. > > Thanks. > > 张铎(Duo Zhang) 于2022年11月4日周五 23:24写道: > > > > If there are no big concerns, I will put up a 2.5.2RC0 next week, with > > hadoop3 binaries, to see if it could meet our users' requirements. > > > > Thanks. > > > > 张铎(Duo Zhang) 于2022年10月31日周一 14:38写道: > > > > > > emmm, after reading the title, seems the word 'quickly' is a bit confusing > > > > > > What I mean here is we can start releasing 2.5.2 earlier than usual, > > > for verify the hadoop3 binaries, not mean the release process for > > > 2.5.2 should be quick. And I even think the release process for 2.5.2 > > > should be longer than before, as we want to change the content of the > > > release. > > > > > > Thanks. > > > > > > 张铎(Duo Zhang) 于2022年10月31日周一 11:58写道: > > > > > > > > Recently I've been working on HBASE-27359 together with some other > > > > folks in the community(Nick, Andrew, Guanghao, etc.). After landing > > > > some important improvements, such as HBASE-27434, HBASE-27442 and > > > > more, now we are ready to make our release scripts able to publish > > > > binaries for hadoop3. > > > > > > > > https://github.com/apache/hbase/pull/4856 > > > > > > > > So here, I wonder if we can quickly put up RCs for 2.5.2, with hadoop3 > > > > binaries and maven artifacts, so our users in the community can test > > > > and use them, report problems. And once we successfully released 2.5.2 > > > > with hadoop3 releases, we can merge the PR for HBASE-27359 back, so > > > > later the release managers can easily publish binaries and maven > > > > artifacts with hadoop3 when making new releases. > > > > > > > > Thanks.