random SMB problems

[jmarchand]

Hi,

My environment
- 6 HPE Apollo 4200 servers with 24 HDD SATA 7.2K 16 Tbytes
- TrueNAS Core :TrueNAS-13.0-U5.3
- ZFS filesystem
- only SMB shares used
- around 30 SMB shares and 150 TBytes per server
- around 150 users per server
- Active-Directory authentication

Since three months, globally all works fine Thanks a lot for TrueNAS

But on two servers, for two different users, on two different shares, I get sometimes problems : the upload of files into a SMB share block.
The problems appear from Windows 10 clients
1/ with the Explorer interface and drag/drop
2/ with CLI and robocopy.

Some files are not completely upload. And the user can do nothing to retry.

The only solution is
a/ directly from the shell on TrueNAS server : remove the file
b/ from the smb client : rename the file

The errors in log.smbd look like

[2023/11/08 17:12:49.806712, 1] ../../source3/smbd/scavenger.c:680(scavenger_timer)
scavenger_timer: Failed to cleanup share modes and byte range locks for file 14510891426233911823:8174187:3436939 open 317801954
[2023/11/08 17:13:15.997243, 1] ../../source3/smbd/scavenger.c:680(scavenger_timer)
scavenger_timer: Failed to cleanup share modes and byte range locks for file 14510891426233911823:8178519:3436943 open 3466303853
[2023/11/08 17:43:28.042692, 1] ../../source3/smbd/scavenger.c:680(scavenger_timer)
scavenger_timer: Failed to cleanup share modes and byte range locks for file 16490179479559229642:3653:148914 open 3335151968

The servers are well-NTP synchronised. I can see with ntpq -p

The files with problems have no characteristic very different with our ten millions of files.

Any idea ?

Best regards,

Joel Marchand

 

[kristyfolks]

Hi,

My environment
- 6 HPE Apollo 4200 servers with 24 HDD SATA 7.2K 16 Tbytes
- TrueNAS Core :TrueNAS-13.0-U5.3
- ZFS filesystem
- only SMB shares used
- around 30 SMB shares and 150 TBytes per server
- around 150 users per server
- Active-Directory authentication

Since three months, globally all works fine Thanks a lot for TrueNAS

But on two servers, for two different users, on two different shares, I get sometimes problems : the upload of files into a SMB share block.
The problems appear from Windows 10 clients
1/ with the Explorer interface and drag/drop
2/ with CLI and robocopy.

Some files are not completely upload. And the user can do nothing to retry.

The only solution is
a/ directly from the shell on TrueNAS server : remove the file
b/ from the smb client : rename the file

The errors in log.smbd look like

[2023/11/08 17:12:49.806712, 1] ../../source3/smbd/scavenger.c:680(scavenger_timer)
scavenger_timer: Failed to cleanup share modes and byte range locks for file 14510891426233911823:8174187:3436939 open 317801954
[2023/11/08 17:13:15.997243, 1] ../../source3/smbd/scavenger.c:680(scavenger_timer)
scavenger_timer: Failed to cleanup share modes and byte range locks for file 14510891426233911823:8178519:3436943 open 3466303853
[2023/11/08 17:43:28.042692, 1] ../../source3/smbd/scavenger.c:680(scavenger_timer)
scavenger_timer: Failed to cleanup share modes and byte range locks for file 16490179479559229642:3653:148914 open 3335151968

The servers are well-NTP synchronised. I can see with ntpq -p

The files with problems have no characteristic very different with our ten millions of files.

Any idea ?

Best regards,

Joel Marchand

Thanks for suggesting this solution. It helps to solve my problem

 

[jmarchand]

Hi,

Thanks for your attention, but I have still no solution and the problems continues to appear.
Any idea ?
Best regards,
Joel