Solved

Gluster scan will not be performed. error protocol error: file name does not match request


Badge +1

Hello Commvault Community :)

 

I’m trying to set-up backup for GlusterFS. I have v11FR22 (11.22.17). On sub-client level I can do “discover” and I’m able “to see” GlusterFS resources, but I can not backup them. During backup I have error message:

 

“glusterfind command could not be executed on host : [localhost] volume : [/mnt/gv06-sec/] due to error : [Function failed during execution. (ERROR_FUNCTION_FAILED.1627), glusterfind command failed. Gluster scan will not be performed. error protocol error: file name does not match request (W32.1627)]”

and in logs files I found:

“Error=[0x8007065B:{CGlusterScanEngine::ProcessWorkUnit(1178)} + {CGlusterScanEngine::ProcessScanVolume(1639)/RunCommand(scp root@localhost:/tmp/GlusterCollect[hostname]_{simpana-device-00-045-1}_4715548.cvf /opt/hds/hds/iDataAgent/jobResults/2/5726/{simpana-device-00-045-1}_0_4715548.cvf) failed} + {CGlusterScanEngine::RunCommand(1728)/W32.1627.(Function failed during execution. (ERROR_FUNCTION_FAILED.1627))-glusterfind command failed. Gluster scan will not be performed. error protocol error: filename does not match request”

 

GlusterFS is configured on RedHat 8, and it is my access node for GlusterFS pseudo-client. I have keyless ssh enabled and it works from command line.

 

I found information: https://bugzilla.redhat.com/show_bug.cgi?id=1722229:

“I think this is caused by the fix for CVE-2019-6111. In this case, the problem are the apostrophes around the filaname.”

 

I check that - there are “apostrophes around the filanames”:

cd /tmp

ls -ltr

-rw-r--r--. 1 root   root     49443 May 11 18:36 'GlusterCollect[hostname_{simpana-device-00-045-1}_4713351.cvf'
-rw-r--r--. 1 root   root     27468 May 11 18:46 'GlusterCollect[hostname]_{simpana-device-00-047-1}_4713333.cvf'
-rw-r--r--. 1 root   root     49307 May 12 12:13 'GlusterCollect[hostname_{simpana-device-00-045-1}_4715548.cvf'

 

And maybe this is a problem… I’m not sure...

 

Does any one faced with this kind of problem and could help how to resolve it ?

Many thanks in advance :)

Dorota

 

icon

Best answer by Mike Struening 8 June 2021, 22:58

Updating and marking resolved :sunglasses:

 

Finding Details:

Filescan.log

---

“Error = [0x8007065B: {CGlusterScanEngine :: ProcessWorkUnit (1178)} + {CGlusterScanEngine :: ProcessScanVolume (1639) / RunCommand (scp root @ localhost: / tmp / GlusterCollect [hostname] _ {simpana-device-00-045-1 } _4715548.cvf /opt/hds/hds/iDataAgent/jobResults/2/5726/[simpana-device-00-045-14015548.cvf) failed} + {CGlusterScanEngine :: RunCommand (1728) /W32.1627. (Function failed during execution. (ERROR_FUNCTION_FAILED.1627)) - glusterfind command failed. Gluster scan will not be performed. error protocol error: filename does not match request "

Solution:

Development provided the following findings:
===
Looks like the customer is backing up the gluster volume  mounted locally on the gluster server itself. We don't need keyless ssh if the gluster server and the data access nodes are the same machine.
Is it possible for the customer to mount the gluster volume using its FQDN instead of localhost and try the backups again ? We will run the command locally instead of running via ssh or copy the gluster output file through scp.
===
After the customer mounted the gluster volume using the FQDN instead of localhost, the backup succeeded.

View original

7 replies

Userlevel 7
Badge +15

@dorota , appreciate the question!  I looked through our database for your error and am not finding anything.  Your bug article definitely seems to be a possible cause, though I can’t confirm.

I’ll tag in some other people for advice.

Userlevel 7
Badge +15

@dorota , just conferred with a few people internally.  Can you open a support case?  We need to investigate thoroughly as we have not seen this before.

Once you do, let me know the case number so I can track it.

Thanks!

Userlevel 4
Badge +6

There has been issues with the use of other special characters in previous service packs but they were resolved. I can find no reference to apostrophes previously but I would assume you are correct that it is the cause of the issue. 

As Mike says please raise a support ticket with all of the necessary logs. 

Badge +1
Thank you for your advice - I'll raise a support ticket :)
Badge +1

 Hi,

 

support ticket is: 210514-118 and we will work on this.

once more - thank you for advice :)

 

Dorothy.

 

Userlevel 7
Badge +15

Thanks!!  I’ll keep an eye on 210514-118.

Userlevel 7
Badge +15

Updating and marking resolved :sunglasses:

 

Finding Details:

Filescan.log

---

“Error = [0x8007065B: {CGlusterScanEngine :: ProcessWorkUnit (1178)} + {CGlusterScanEngine :: ProcessScanVolume (1639) / RunCommand (scp root @ localhost: / tmp / GlusterCollect [hostname] _ {simpana-device-00-045-1 } _4715548.cvf /opt/hds/hds/iDataAgent/jobResults/2/5726/[simpana-device-00-045-14015548.cvf) failed} + {CGlusterScanEngine :: RunCommand (1728) /W32.1627. (Function failed during execution. (ERROR_FUNCTION_FAILED.1627)) - glusterfind command failed. Gluster scan will not be performed. error protocol error: filename does not match request "

Solution:

Development provided the following findings:
===
Looks like the customer is backing up the gluster volume  mounted locally on the gluster server itself. We don't need keyless ssh if the gluster server and the data access nodes are the same machine.
Is it possible for the customer to mount the gluster volume using its FQDN instead of localhost and try the backups again ? We will run the command locally instead of running via ssh or copy the gluster output file through scp.
===
After the customer mounted the gluster volume using the FQDN instead of localhost, the backup succeeded.

Reply