'du' and 'df' tools report different space utilization
Recently in my quater rack Exadata system (2 nodes, 11202 database on top of 5.5 OEL), /u01 had 0 free disk space. After I identified a 65 GB trace file and removed it, "df -h" still show 0 free disk space, but when I did "du -sh /u01", it show only 24 GB disk space was used, and the /u01 has a capacity of 89 GB. After some research, I found this note: 'du' and 'df' tools report different space utilization [ID 457444.1]. By following the instruction in this note, I fixed this different space utilization report problem. Here is how:
1 run the lsof command (/usr/sbin/lsof | grep deleted) as root to identify the holding process:
[root@node02 scripts]# lsof | grep deleted
oracle 10373 oracle 4w REG 253,2 10538879 13041668 /u01/app/11.2.0/grid/log/thordb02/agent/ohasd/oraagent_oracle/oraagent_oracle.l10 (deleted)
oracle 10375 oracle 4w REG 253,2 10538879 13041668 /u01/app/11.2.0/grid/log/thordb02/agent/crsd/oraagent_oracle/oraagent_oracle.l10 (deleted)
oracle 11028 oracle 4w REG 253,2 10532777 3817484 /u01/app/11.2.0/grid/log/thordb02/agent/crsd/oraagent_oracle/oraagent_oracle.l10 (deleted)
oracle 11028 oracle 38w REG 253,2 69055365120 7344857 /u01/app/oracle/diag/rdbms/thor/thor2/trace/thor2_smon_11028.trc (deleted)
Note:
- the 7th column in the output denotes the size of deleted files (in bytes).
- The 9th column denotes which file remains held open.
- The 1st and second columns denotes the process and pid that still holds the open file descriptor.
2 kill the "11028" process identified from the above
#kill -9 11028
3 redo "df -h" and "du -sh /u01" to see if they report same space utilization:
[root@node2 scripts]# df -h
Filesystem Size Used Avail Use% Mounted on
/dev/mapper/VGExaDb-LVDbSys1
30G 5.2G 23G 19% /
/dev/sda1 124M 16M 102M 14% /boot
/dev/mapper/VGExaDb-LVDbOra1
99G 30G 65G 32% /u01
tmpfs 81G 196M 81G 1% /dev/shm
[root@node2 scripts]# du -sh /u01
29G /u01
Problem solved.
0 Comments:
Post a Comment
Subscribe to Post Comments [Atom]
<< Home