ISE /opt /tmp full

[AK002]

Hi Team,

Getting Alert in ISE, How can we proceed further or can we remove backups and other old files. 

Version : 2.1.0.474 patch 2

Ise01/admin# show disks

Internal filesystems:
/ : 18% used ( 2417172 of 14987616)
/dev : 0% used ( 0 of 8124320)
/dev/shm : 0% used ( 0 of 8134236)
/run : 1% used ( 8956 of 8134236)
/sys/fs/cgroup : 0% used ( 0 of 8134236)
/boot : 23% used ( 103720 of 487634)
/storedconfig : 2% used ( 1582 of 95054)
/tmp : 100% used ( 1966668 of 1983056)
/opt : 93% used ( 158812448 of 180485852)
/run/user/440 : 0% used ( 0 of 1626848)
/run/user/301 : 0% used ( 0 of 1626848)
/run/user/0 : 0% used ( 0 of 1626848)
/run/user/304 : 0% used ( 0 of 1626848)
/run/user/303 : 0% used ( 0 of 1626848)
/run/user/322 : 0% used ( 0 of 1626848)
/run/user/321 : 0% used ( 0 of 1626848)
warning - /tmp is 100% used (1966668 of 1983056)
warning - /opt is 93% used (158812448 of 180485852)

Regards,

AK

[marce1000]

 - You will have to contact TAC and ask for the (a) root patch ; which will allow you to remove unwanted files manually (or have it done by TAC).

M.

[AK002]

--So we need to raise a TAC case in order to remove those files from directories. There is no other way to cleanup by admins manually. 

AK

[marce1000]

 - Yes, and that is a very good point;  I have always felt that if you use open source as the underlying OS ; you should have the customers  benefit from it too (taken into account open-source philosophy) . In Prime , one can become root. Can't understand why  it wasn't made that way in  ISE.

M.

[AK002]

Yes in PI we can do in root but ISE they keeping in secret. :)

[MDBee]

If the /opt directory usage is too high then configuration backups will fail. In my case the /opt directory was at 76% usage. To reduce the size of it I had a TAC case with Cisco. They attempted to reduce space with root access with no success. The fix when working with cisco was the below (this does not require root access).

1. application configure ise
2. Type "1" to select the "Reset M&T Database".

This took 10-15 minutes to complete but after it was complete the /opt directory was then down to 24% usage. 

Note: It doesn't work on the primary node as the database is being used by the active session. You need to promote the secondary node to the primary and ran the "Reset M&T Database" on the ISE node that is now the secondary one. If this still doesn't help further Cisco TAC help may be required.