2024-11-05T01:07:27.656Z

<Xiubo Li> @Venky Shankar @Markuze There is high priority bug reported by Luis in kclient, could you guys follow it ? Currently I do not have the develop env and it's hard for me to go through the source code and also I'm a bit busy with learning new things every day, so not sure I can push it in time.

`[RFC PATCH v2] ceph: ceph: fix out-of-bound array access when doing a file read`

Thanks.

2024-11-05T04:44:15.600Z

<Anthony D'Atri> A perennial problem :-/ Going back to Brad Cox who routinely filled up the homedir filesystem.  

2024-11-05T05:51:18.875Z

<Venky Shankar> Thanks for letting us know @Xiubo Li 👍

2024-11-05T05:52:13.721Z

<Venky Shankar> @Milind Changire Wanted to get your attention on <https://tracker.ceph.com/issues/68747#note-3> (cc @Patrick Donnelly)

2024-11-05T06:25:34.254Z

<Neeraj Pratap Singh> @vshankar Can u pls have  a quick look at <https://github.com/ceph/ceph/pull/59019#issuecomment-2456307074> and <https://github.com/ceph/ceph/pull/56634#issuecomment-2456329265> (short comments)

2024-11-05T06:25:57.721Z

<Neeraj Pratap Singh> @Venky Shankar Can u pls have  a quick look at <https://github.com/ceph/ceph/pull/59019#issuecomment-2456307074> and <https://github.com/ceph/ceph/pull/56634#issuecomment-2456329265> (short comments)

2024-11-05T06:25:59.220Z

<Venky Shankar> Noted.

2024-11-05T09:04:17.653Z

<Igor Golikov> Hi, does the `vstart` cluster runs completely in memory? When `stop.sh` is called and finished, can I assume that no previous state of the cluster data/metadata has been left and I can just run it again?

2024-11-05T09:06:26.690Z

<jcollin> rm -Rf out/

2024-11-05T09:24:03.035Z

<Venky Shankar> @Igor Golikov I does have persistent storage, so next time you invoke vstart.sh without `-n` flag, it would reuse existing ceph.conf

2024-11-05T10:53:36.596Z

<Markuze> @Xiubo Li, wellcome back 🙂
.Where is this bug? Is it on a mailing list?

2024-11-05T10:59:04.744Z

<Markuze> Oh I see it in the ibm email.

2024-11-05T10:59:35.980Z

<henrix> it's on the mailing-list: https://lore.kernel.org/all/20240905135700.16394-1-luis.henriques@linux.dev/

2024-11-05T11:02:10.225Z

/me <henrix> is on IRC and can't ping/reference slack nicks :-(

2024-11-05T11:36:04.077Z

<Xiubo Li> Thanks @Markuze Yeah，it's in the mail list, I CCed you

2024-11-05T23:22:49.511Z

<Seungmin Kim> Is there ANY way to restrict the user or cluster IOPS in CephFS from Rook or Ceph? This is to prevent the metadata servers from flooding and breaking down. Had to recently do [https://silvenga.com/posts/notes-on-cephfs-metadata-recovery/](https://silvenga.com/posts/notes-on-cephfs-metadata-recovery/) because of a flood in the metadata server.