@@ -222,20 +222,27 @@ problems first (:doc:`../../rados/troubleshooting/index`).
222222The MDS
223223=======
224224
225- If an operation is hung inside the MDS, it will eventually show up in ``ceph health ``,
226- identifying "slow requests are blocked". It may also identify clients as
227- "failing to respond" or misbehaving in other ways. If the MDS identifies
228- specific clients as misbehaving, you should investigate why they are doing so.
225+ Run the ``ceph health `` command. Any operation that is hung in the MDS is
226+ indicated by the ``slow requests are blocked `` message.
229227
230- Generally it will be the result of
228+ Messages that read ``failing to respond `` indicate that a client is failing to
229+ respond.
231230
232- #. Overloading the system (if you have extra RAM, increase the
233- "mds cache memory limit" config from its default 1GiB; having a larger active
234- file set than your MDS cache is the #1 cause of this!).
231+ The following list details potential causes of hung operations:
235232
236- #. Running an older (misbehaving) client.
233+ #. The system is overloaded. The most likely cause of system overload is an
234+ active file set that is larger than the MDS cache.
235+
236+ If you have extra RAM, increase the ``mds_cache_memory_limit ``. The specific
237+ tunable ``mds_cache_memory_limit `` is discussed in the :ref: `MDS Cache
238+ Size<cephfs_cache_configuration_mds_cache_memory_limit>`. Read the :ref: `MDS
239+ Cache Configuration<cephfs_mds_cache_configuration>` section in full before
240+ making any alterations to the ``mds_cache_memory_limit `` tunable.
237241
238- #. Underlying RADOS issues.
242+ #. There is an older (misbehaving) client.
243+
244+ #. There are underlying RADOS issues. See :ref: `The RADOS troubleshooting
245+ documentation<rados_troubleshooting>`.
239246
240247Otherwise, you have probably discovered a new bug and should report it to
241248the developers!
0 commit comments