Skip to content

Commit 16d593e

Browse files
authored
Merge pull request ceph#64881 from zdover23/wip-doc-2025-08-07-cephfs-troubleshooting-3
doc/cephfs: edit troubleshooting.rst Reviewed-by: Anthony D'Atri <[email protected]>
2 parents dfdc970 + 57e7be7 commit 16d593e

File tree

1 file changed

+23
-11
lines changed

1 file changed

+23
-11
lines changed

doc/cephfs/troubleshooting.rst

Lines changed: 23 additions & 11 deletions
Original file line numberDiff line numberDiff line change
@@ -5,21 +5,33 @@
55
Slow/stuck operations
66
=====================
77

8-
If you are experiencing apparent hung operations, the first task is to identify
9-
where the problem is occurring: in the client, the MDS, or the network connecting
10-
them. Start by looking to see if either side has stuck operations
11-
(:ref:`slow_requests`, below), and narrow it down from there.
8+
Sometimes CephFS operations hang. The first step in troubleshooting them is to
9+
locate the problem causing the operations to hang. Problems present in three
10+
places:
1211

13-
We can get hints about what's going on by dumping the MDS cache ::
12+
#. in the client
13+
#. in the MDS
14+
#. in the network that connects the client to the MDS
1415

15-
ceph daemon mds.<name> dump cache /tmp/dump.txt
16+
First, use the procedure in :ref:`slow_requests` to determine if the client has
17+
stuck operations or the MDS has stuck operations.
1618

17-
.. note:: The file `dump.txt` is on the machine executing the MDS and for systemd
18-
controlled MDS services, this is in a tmpfs in the MDS container.
19-
Use `nsenter(1)` to locate `dump.txt` or specify another system-wide path.
19+
Dump the MDS cache. The contents of the MDS cache will be used to diagnose the
20+
nature of the problem. Run the following command to dump the MDS cache:
2021

21-
If high logging levels are set on the MDS, that will almost certainly hold the
22-
information we need to diagnose and solve the issue.
22+
.. prompt:: bash #
23+
24+
ceph daemon mds.<name> dump cache /tmp/dump.txt
25+
26+
.. note:: MDS services that are not controlled by systemd dump the file
27+
``dump.txt`` to the machine that runs the MDS. MDS services that are
28+
controlled by systemd dump the file ``dump.txt`` to a tmpfs in the MDS
29+
container. Use `nsenter(1)` to locate ``dump.txt`` or specify another
30+
system-wide path.
31+
32+
If high logging levels have been set on the MDS, ``dump.txt`` can be expected
33+
to hold the information needed to diagnose and solve the issue causing the
34+
CephFS operations to hang.
2335

2436
Stuck during recovery
2537
=====================

0 commit comments

Comments
 (0)