Disk Bottleneck & Network Bottleneck

[Ishwar_1]

Hi Gurus,

I am getting an message of Disk bottleneck probability to 100% similar with network from glance. I have posted the output of the dmesg which shows that 2 filesystem are full but when i view it with bdf it is less then 50%.

output for dmesg
paging to/from NFS server avfermion
file system is /redologs
Synchronous Page I/O error occurred while paging to/from NFS server avfermion
file system is /redologs
Synchronous Page I/O error occurred while paging to/from NFS server avfermion
vxfs: NOTICE: msgcnt 29 mesg 001: V-2-1: vx_nospace - /dev/vg02/lvol13 file system full (512 block extent)
vxfs: NOTICE: msgcnt 31 mesg 001: V-2-1: vx_nospace - /dev/vg02/lvol13 file system full (1024 block extent)
Synchronous Page I/O error occurred while paging to/from NFS server avfermion
file system is /redologs
vxfs: NOTICE: msgcnt 85 mesg 001: V-2-1: vx_nospace - /dev/vg02/lvol13 file system full (1024 block extent)
vxfs: NOTICE: msgcnt 87 mesg 001: V-2-1: vx_nospace - /dev/vg02/lvol13 file system full (2048 block extent)
Synchronous Page I/O error occurred while paging to/from NFS server avfermion
file system is /redologs
vxfs: NOTICE: msgcnt 93 mesg 001: V-2-1: vx_nospace - /dev/vg02/lvol13 file system full (1024 block extent)
Synchronous Page I/O error occurred while paging to/from NFS server avfermion
file system is /redologs
Synchronous Synchronous Page I/O error occurred while paging to/from NFS server avfermion
file system is /redologs
Page I/O error occurred while paging to/from NFS server avfermion
file system is /redologs
Synchronous Page I/O error occurred while paging to/from NFS server avgraviton
file system is /sapshareOLD
NFS server avfermion not responding still trying
NFS server avfermion ok
vxfs: NOTICE: msgcnt 95 mesg 001: V-2-1: vx_nospace - /dev/vg01/lvol10 file system full (8 block extent)
vxfs: NOTICE: msgcnt 97 mesg 001: V-2-1: vx_nospace - /dev/vg01/lvol10 file system full (8 block extent)
vxfs: NOTICE: msgcnt 99 mesg 001: V-2-1: vx_nospace - /dev/vg01/lvol10 file system full (8 block extent)
vxfs: NOTICE: msgcnt 101 mesg 001: V-2-1: vx_nospace - /dev/vg01/lvol10 file system full (32 block extent)
vxfs: NOTICE: msgcnt 103 mesg 001: V-2-1: vx_nospace - /dev/vg01/lvol10 file system full (32 block extent)

Regards
Ishwar

[Steven E. Protter]

Shalom Ishwar,

1) bdf on the system. Clean up the full file system problems. See if your other problems go away.

2) Check statistics when the system is running normally and I/O isn't backing up, waiting to write to disks that might be full.

SEP

[Ishwar_1]

Steven,

I have attached the output of the bdf which shows the 2 logical volume lvol10 and lvol13 are having space less then the 45%

Regards

[rick jones]

I believe a disc bottleneck warning in glance comes from the I/O rates to the discs, not how full the filesystems on them might be.

[Kapil Jha]

I agree with Rick, disk bottle neck in Glance is actually the IO on the disk, so you need to check which disk and/or which LV is having high IO,

regarding file system full msg in messages file
/oracle/PRD/oraarch and /eu_diveline
Arch folder is probable to get full.
Are you sure that you getting same error again?

for bottle neck you have to see the which disk is having IO.

BR,
Kapil+

[Dennis Handly]

>shows the 2 logical volume lvol10 and lvol13 are having space less then the 45%

Instead of looking at old stale dmesg data, look at syslog.log.

[Rita C Workman]

I'm agree. Do you really have a problem?

You printed dmesg - old info
You printed the Alarm screen on Glance - may be old info, hence why they have a Clear option in the toolbar.

Use syslog, use the clear button in Glance to ensure you're not looking at old info. Another good one to clear first is vmstat with a vmstat -z, so when you see a value in 'paged-out' you know it's an issue and not just a past occurance.

Confirm that you are sending 'current' conditions to be addressed.

Kindest regards,
Rita

[Ishwar_1]

Hi Gurus,

I have checked syslog.log their was message related to Filesystem but i am not able to figure it out. But currently i am getting similar alert for the /dev/vg01/lvol4 which is 61%. I have attached the error log which i am getting today from syslog.log.
------------------------------------------
As far my research i found that

Disk bottleneck message is related to disk access performance, not for available disk space. Usually its due to application overloading on an individual disk/disk subsystem it may also be due to faster disk utilization I/O operation.

Network bottleneck probability message is related to network traffic We can check the Network interface statistics for any collision and any other erroneous packets, choking the network.

Regards
Ishwar