1 Reply Latest reply on Mar 4, 2013 9:15 AM by alvaromiranda

    Oracle Linux ES 5.5 - Local Hard Disk Free space release issue

      We are running Oracle E-Business Suite Release 12 (12.1.3) / Production Instance on Oracle Linux 5.x

      Oracle Application Two Tier Environment:
      Node 1 – (Oracle Linux) – Database
      Node 2 – (Oracle Linux) – All Application Services

      Operating System Details:
      [root@nfs-prod-app spool]# cat /etc/redhat-release
      Red Hat Enterprise Linux Server release 5.5 (Tikanga)
      [root@nfs-prod-app spool]#

      The total /u01 mount point size is 105G where /u01/applprod/apps and /u01/applprod/inst instance directory exsits.
      we have two directories
      /u01/applprod - 57 GB
      /u01/appldisc - 22 GB

      Client is installed other monitoring tools under /u01 mount point it occupied another 20GB in this mount point. Later they have reliazed and delete all the monitoring tool files.
      when we do df –h it still shows only 2.7 GB available and 97GB is occupied. The free space is not claiming after deleting the monitoring tool files.

      Actually it should be 25GB availblae but it shows only 2.7GB is available.
      Because of this insufficient space on /u01 the oracle application services are malfunctioning.

      Please help how to release the free space under /u01 mount point
        • 1. Re: Oracle Linux ES 5.5 - Local Hard Disk Free space release issue
          what is the filesystem you are using there?

          mount | grep /u01

          if the filesystem is OCFS2, in old versions there is a bug in some cases the space won't be released and will fell undersomething called orphan files.

          if that's the ase, you should stop the services, umount the filesystem on alll the nodes on the cluster, and run fsck

          If the filesystem is not ocfs2, then maybe you still have process using those files/directories and the data is marked as deleted, but will be released once those process finish.

          as root run

          lsof | grep delete

          and check if there are files marked as deleted under the path that match where this monitoring tool was.
          If yes, then try to kill those process and check the size again.

          If none of the 2 cases address your problem, please provide more information

          df -h