Can we delete or move dev_enqwork file, when system is up & running

[Former Member]

Hello Experts,

Can we delete or remove the dev_enqwork file while system is up and running.

Location of this file is /usr/sap/SID/SCS**/work.

As the size of this file has been reached upto 9 GB and we are running out of space.

Thanks.

[csaba_goetz · Accepted solution]

Hello Anand,

Yes , you can.

First of all save the last n lines or the complete dev_enqwork in another location - to be on the safe side, not to loose any information what could be relevant, e.g. with command:

tail -1000 /usr/sap/SID/SCS**/work/dev_enqwork > another_folder/dev_enqwork.bck

Afterwards empty dev_enqwork, e.g. this way

echo cut > /usr/sap/SID/SCS**/work/dev_enqwork

date > /usr/sap/SID/SCS**/work/dev_enqwork

After this you can check that its size is much smaller:

ls -al /usr/sap/SID/SCS**/work/dev_enqwork

Best regards,

Adam

[Former Member · Accepted solution]

Thanks Adam...it helps.

[Former Member]

Something must be wrong, did anybody increase the trace level for the enqeue process?

You can delete the file manually, but you should better reset the files properly. For standalone central services instances, this can be done with the ensmon tool.

sidadm> cdpro
sidadm> ensmon pf=SID_SCS0x_server
Try to connect to host server service sapdp00
Enqueue Server monitor main menu
================================
   1: Dummy request
   2: Get replication information
   3: Set/get trace status information on/from the enque server
   4: Reload Enqueue Server ACL file (simple request)
   5: get Enqueue Server ACL status
   6: dump Enqueue Server ACL
   7: Get information about all enqueue server threads
   8: Get information about enqueue memory
   9: dump Enqueue Parameter
   q: quit
   h: help
 ==>3
Trace action to do in the enqueue server:
1: raise, 2: lower, 3: get status, 4: set status, 5: reset files (default=3):
==> 5
Area the action shall be limited to (default: EN):
  EN: enqueue server
  THIO<n>: I/O thread n (no <n> specified: all I/O threads)
  THEN: enqueue- (worker-) thread
  THREP: replication thread
  THAD: ADM thread
==> en
Sending request to server ...
Succeeded to reset the trace file(s)
of the enqueue server

Cheers Michael