Latency with filesystems ending with _netdev attached to them
Users are complaining about super slow response times when they are copying data from Oracle server over to the storage device where these filesystems sit. Checked Storage and Network and both are clean, no issues in the logs.
I have filesystems listed below that show _netdev at the end of each line in /etc/fstab /dev/helixoravg/oracle_data2_lv /oracle_data2 xfs _netdev 0 0 /dev/helixoravg/oracle_redo2_lv /oracle_redo2 xfs _netdev 0 0 /dev/orabackupvg/oracle_archive2_lv /oracle_archive2 xfs _netdev 0 0 /dev/orabackupvg/oracle_backup2_lv /oracle_backup2 xfs _netdev 0 0 I was told these should not be transferring over the network and using ISCI only. Anyone know how to fix this? |
What distro and version are you using?
Are those logical volumes local on the server or remote? If local then _netdev would probably not be correct as a mount option. It looks like you are missing the options field in those fstab entries. Since you posted without using the code tags I can see no space between xfs and _netdev. Is it "xfs_netdev" or "xfs _netdev"? Can't tell. The lines would be much more readable if they looked like this Code:
/dev/mapper/fedora-root / ext4 defaults 1 1 /dev/mapper/fedora-root / ext4 defaults 1 1 That is the same line posted with code tags and without. |
computersavvy, just hit the "Quote" button on the OP - you'll see the full layout as it was posted.
The mount manpage states that _netdev is only to wait for the network to come up before the mount is attempted. Can't see why it would affect operations after that. |
Quote:
Right and the above is ambiguous and implies data is being transferred across the network. I asked for additional information as well as suggesting the use of code tags, although I did misunderstand the man page. |
Quote:
To fix performance problems like slow storage, I'd do something like this:
|
Awesome information. Thank you all!!
|
All times are GMT -5. The time now is 06:13 AM. |