[Gluster-users] ls slow on a GlusterFS mounted filesystem

[Milos Cuculovic]

We have two replicated GlusterFS servers. When trying to run ls on any of the GlusterFS mounted directories, it takes long time to process.
For example, on a irectory with 10 files and 10 subdirectories, it takes 5sec to do the ls.

Here is the volume config info:

Volume Name: storage2
Type: Replicate
Volume ID: 8af504b7-8a95-463b-8578-9a969ae060ff
Status: Started
Snapshot Count: 0
Number of Bricks: 1 x 2 = 2
Transport-type: tcp
Bricks:
Brick1: storage3:/data/data-cluster
Brick2: storage4:/data/data-cluster
Options Reconfigured:
storage.build-pgfid: on
cluster.readdir-optimize: off
storage.fips-mode-rchecksum: on
client.event-threads: 12
server.event-threads: 12
nfs.disable: on
transport.address-family: inet
auth.allow: 10.10.0.*
features.cache-invalidation: on
features.cache-invalidation-timeout: 600
performance.stat-prefetch: on
performance.readdir-ahead: on
performance.cache-refresh-timeout: 1
network.compression.compression-level: -1
network.compression: off
cluster.min-free-disk: 2%
performance.cache-size: 1GB
features.bitrot: on
features.scrub: Active
performance.client-io-threads: on
features.scrub-freq: monthly
features.scrub-throttle: normal

Any idea on how to improve this?

Looking forward to hearing from you.

Milos