zpool iostat value bigger than fsstat
Dear ZFS experts ..
We have 'zpool iostat' and 'fsstat' commands running in our Oracle Database 11G server. We are monitoring the I/O load rite now.
What I found interesting is "zpool iostat" shows much more READ IOPS than "fsstat". What does it mean ? Is it an indication of performance degradation ? That ZFS requires a lot more I/O reading from disks for its internal operation rather than serving the application (Oracle Database) ?
p004 is one of our ZFS filesystem that stores Oracle datafiles.
Output from "zpool iostat 1" :
capacity operations bandwidth
pool alloc free read write read write
p004 2.95T 2.49T 4.25K 35 537M 4.48M
We have 'zpool iostat' and 'fsstat' commands running in our Oracle Database 11G server. We are monitoring the I/O load rite now.
What I found interesting is "zpool iostat" shows much more READ IOPS than "fsstat". What does it mean ? Is it an indication of performance degradation ? That ZFS requires a lot more I/O reading from disks for its internal operation rather than serving the application (Oracle Database) ?
p004 is one of our ZFS filesystem that stores Oracle datafiles.
Output from "zpool iostat 1" :
capacity operations bandwidth
pool alloc free read write read write
p004 2.95T 2.49T 4.25K 35 537M 4.48M
0