OCS/ODF How to identify the PV cephfs associated to a damaged file reported by mds

Solution Verified - Updated -

Issue

The mds pod crashed with this mds logs just before the crash:

2022-11-03T13:07:17.979984524Z debug     -2> 2022-11-03T13:07:17.909+0000 7fac1974f700  4 mds.0.server handle_client_request client_request(client.32036757:6395 unlink #0x100068b2b10/000000010000000300000053 2022-11-03T13:07:17.909539+0000 caller_uid=1001530000, caller_gid=0{1001530000,}) v4
2022-11-03T13:07:17.979993835Z debug     -1> 2022-11-03T13:07:17.912+0000 7fac1974f700 -1 /builddir/build/BUILD/ceph-16.2.7/src/mds/Server.cc: In function 'void Server::_unlink_local(MDRequestRef&, CDentry*, CDentry*)' thread 7fac1974f700 time 2022-11-03T13:07:17.911262+0000
  • Here mds pod was trying to do a unlink of file "#0x100068b2b10/000000010000000300000053" at "2022-11-03T13:07:17" UTC so we suspect some metadata of that file is damaged.
  • The name of the file is "000000010000000300000053"
  • The name of the directory where this file is located is unknown, we only know the inode number "0x100068b2b10" (note: this value reported by mds is an hexadecimal number) of that directory.

The object of this document is to find out the PV name where this file is located.

Environment

Red Hat OpenShift Container Storage (RHOCS) v4.x
Red Hat OpenShift Data Foundations (RHODF) x4.x

Subscriber exclusive content

A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.

Current Customers and Partners

Log in for full access

Log In

New to Red Hat?

Learn more about Red Hat subscriptions

Using a Red Hat product through a public cloud?

How to access this content