cancel
Showing results for 
Search instead for 
Did you mean: 

INTERMITTENTLY Error 22 for write/read access to a file on sapmnt

former_member211576
Contributor
0 Kudos

Hi experts,

  We run our SAP on SQL Server with Windows Server Failover Cluster and I migrated one DC(qrdcsapdc2) to another DC(qrcsapdc7) a couple weeks ago. We found a few jobs failed from 9/8 2:00am ~ 9/8 4:00am in SM37. I also found there are many error 22 for write/read access to a file on sapmnt in SM21.

  1. 1. The first weird part is the timeframe is very regular. Its occurred only from 2:00am ~ 4:00am(9/6 2:30am happened, too)
  2. 2. The second weird part is we have 7 Additional Application Servers(tccap51 ~ tccap57) and it occurred only on 3 AASs(tccap53, tccap54, tccap55)
  3. 3. By the way, I have disable 8dot3name(fsutil 8dot3name set S: 1) according to note 662452 on 9/6 but I have not restarted OS yet.

File = \\tccprd1\sapmnt\TCP\SYS\global\218JOBLG\0001X01450800X50446

Cannot close the file of a TemSe object (error no 22)

Possible causes & solutions?

  1. 1. Something wrong on Active Directory(DC) authentication issues and I prepare to restart all DCs, DBs and AASs to get new authentication tickets.
  2. 2. Intermittent network packet loss?
  • Replace with new network cables and switch ports
  1. 3. Others….

Please help. Any comments are appreciated.

Accepted Solutions (1)

Accepted Solutions (1)

S_Sriram
Active Contributor
0 Kudos

Hi Lee.

1. check the disk space in MS Sql DB, system and usr folers.

2. During off time just do the full system restart and the check the SAP system.

3. Check the SAP note 48400  for the Temse db error

BR

SS

former_member211576
Contributor
0 Kudos

Hi SS,

  1. There are a lot of free disk space for SQL DB & sapmnt.

  2. I might restart because I have no clues at all right now.

  3. I have read note 1909093 and 48400 and "Yes" RSPO1043 & RSTS0043 run periodically.

S_Sriram
Active Contributor
0 Kudos

Hi Lee.

Thanks for your information.

1. During the error time you have to check the Temse  db any spool request went failed or not.

2. During this error your cluster group failover to another node? may be because of this you will get this error. Could you check the failover cluster event log any  resources fail over to other node?

BR

SS

former_member211576
Contributor
0 Kudos

Hi SS,

  1. OK. but why did it fail all of a sudden at 2:00am? Did AAS can't access sapmnt at that point in time? and why?

  2. No. There are 7 APP servers and 21 SAP AASs(SAP instances) and there are always  a lot of jobs running. Even inside the problematic AASs(tccap53, tccap54, tccap55), there are still a lot of jobs ran successfully. Not all jobs on those AASs failed.

0 Kudos

Hello Denis,

which Windows Version do you use for your Failover Cluster? If Windows Server 2012 or higher, do you have enabled the CA flag (Continuous Availabilty) of the sapmnt share?

Error 22 points to a temporary problem with a system driver, either on the host which tries to access the remote file share or on the host, where the file share is located (in your case the Failover Cluster).

Are you able to reproduce the issue? Or is it "just sporadic"?

Is it related to specific application servers or any?

Best regards,

Kalle

former_member211576
Contributor
0 Kudos

Hi Karl,

  Thanks for your reply. I have created an incident to SAP and the problem is solved.

---

Hi <xxx>,
  Thanks for your response. It is very useful. I put the SAP directories(sapmnt) together with our secondary database and our backup destination volume in one single SAN storage. I can find average disk sec/write is 47 mini-seconds and peak disk sec/write is 992 mini-seconds(please check the attachment). I think it is the source of the problem. I will move backup destination volume to other storage. By the way, we don't install anti-virus software on our DB server.

----

Hi Customer,

File = \\tccprd1\sapmnt\TCP\SYS\global\218JOBLG\0001X01450800X50446
Cannot close the file of a TemSe object (error no 22)

Normally it is not a permission problem unless you have it in all the  jobs.
OS error 22 can be anything. Temporary unavailable disk (due to  backup or interfering antivirus scanner),
high load problem on the NTFS  filesystem or Network,  Windows OS resource bottleneck at that point in time and more.

I have seen several issues of this kind in the past which  occured usually in high load situations.
And in many cases it was the anti-virus scanner. For this kind of problem we recommend to make sure antivirus scanner is configured to
EXCLUDE the realtime scan  to the SAP  directories (\usr\sap folders).

Answers (0)