Why is the 'glfs_fusenoti' Gluster thread showing more than 90% of IO in the output of 'iotop'?
Issue
-
The command iotop shows a very high IO in a gluster mount point, specially in the
glfs_fusenotithread. However, there are no read and writes reported:TID PRIO USER DISK READ DISK WRITE SWAPIN IO COMMAND 53257 be/4 root 0.00 B/s 0.00 B/s 0.00 % 97.16 % glusterfs --acl --use-readdirp=no --volfile-id=/test-vol /test-vol [glfs_fusenoti]Why is this occurring?
-
Important: Please note that this document applies to the specific case where the the high IO usage points to the
glfs_fusenotithread. If any other thread is responsible for this high IO usage, then the steps in this article are not applicable.
Environment
- Red Hat Gluster Storage version 3.5.4 and below.
- Red Hat Openshift Container Storage version 3.11.7 and below.
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.
