Skip to content

Commit b627eb4

Browse files
committed
doc/cephfs: edit troubleshooting.rst
Edit the section "Expediting MDS Journal Trim" in doc/cephfs/troubleshooting.rst. Signed-off-by: Zac Dover <[email protected]>
1 parent a78281e commit b627eb4

File tree

1 file changed

+15
-8
lines changed

1 file changed

+15
-8
lines changed

doc/cephfs/troubleshooting.rst

Lines changed: 15 additions & 8 deletions
Original file line numberDiff line numberDiff line change
@@ -86,6 +86,7 @@ Replay completes when the ``journal_read_pos`` reaches the
8686
``journal_write_pos``. The write position does not change during replay. Track
8787
the progression of the read position to compute the expected time to complete.
8888

89+
.. _cephfs_troubleshooting_avoiding_recovery_roadblocks:
8990

9091
Avoiding recovery roadblocks
9192
----------------------------
@@ -197,17 +198,23 @@ Do the following when restoring your file system:
197198
Expediting MDS journal trim
198199
===========================
199200

200-
If your MDS journal grew too large (maybe your MDS was stuck in up:replay for a
201-
long time!), you will want to have the MDS trim its journal more frequently.
202-
You will know the journal is too large because of ``MDS_HEALTH_TRIM`` warnings.
201+
``MDS_HEALTH_TRIM`` warnings indicate that the MDS journal has grown too large.
202+
When the MDS journal has grown too large, use the ``mds_tick_interval`` tunable
203+
to modify the "MDS tick interval". The "tick" interval drives various upkeep
204+
activities in the MDS, and modifying the interval will decrease the size of the
205+
MDS journal by ensuring that it is trimmed more frequently.
203206

204-
The main tunable available to do this is to modify the MDS tick interval. The
205-
"tick" interval drives several upkeep activities in the MDS. It is strongly
206-
recommended no significant file system load be present when modifying this tick
207-
interval. This setting only affects an MDS in ``up:active``. The MDS does not
207+
Make sure that there is no significant file-system load present when modifying
208+
``mds_tick_interval``. See
209+
:ref:`cephfs_troubleshooting_avoiding_recovery_roadblocks` for ways to reduce
210+
load on the CephFS.
211+
212+
This setting affects only MDSes in the ``up:active`` state. The MDS does not
208213
trim its journal during recovery.
209214

210-
.. code:: bash
215+
Run the following command to modify the ``mds_tick_interval`` tunable:
216+
217+
.. prompt:: bash #
211218

212219
ceph config set mds mds_tick_interval 2
213220

0 commit comments

Comments
 (0)