[Gluster-users] "Transport endpoint is not connected" error + long list of files to be healed

[Mauro Tridici]

Dear All,

our GlusterFS filesystem was showing some problem during some simple users actions (for example, during directory or file creation).

> mkdir -p test
> mkdir: impossibile creare la directory `test': Transport endpoint is not connected


After received some users notification, I investigated about the issue and I detected that 3 bricks (each one in a separate gluster servers) were down.
So, I forced the bricks to be up using “gluster vol start tier force” and bricks come back successfully. All the bricks are up.

Anyway, I see from “gluster vol status” command output that also 2 self-heal daemons were down and I had to restart daemons to fix the problem.
Now, everything seems to be ok watching the output of “gluster vol status” and I can create a test directory on the file system.

But, during the last check made using “gluster volume heal tier2 info”, I saw a long list of files and directories that need to be healed.
The list is very long and the command output is still going ahead on my terminal.

What I can do to fix this issue? Does the self-heal feature fix automatically each files that need to be healed?
Could you please help me to understand what I need to do in this case?

You can find below some information about our GlusterFS configuration:

Volume Name: tier2
Type: Distributed-Disperse
Volume ID: a28d88c5-3295-4e35-98d4-210b3af9358c
Status: Started
Snapshot Count: 0
Number of Bricks: 12 x (4 + 2) = 72
Transport-type: tcp

Thank you in advance.
Regards,
Mauro
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.gluster.org/pipermail/gluster-users/attachments/20191113/fd4f0f4e/attachment.html>