Getting errors from Cron jobs

[Ted Ellis_2]

we have an application running that has several monitor scripts in crontabs. They historically have run fine. As of yesterday, we are getting messages like below... any ideas?

/export/home/prod//config.guess[34]: /dev/null: cannot create
(Unable to guess system type)
/export/home/prod/smc_smc/scripts/Monitor/CheckOutboundSmcInts.ss[1886351988]: cannot make pipe
qdump: Could not open file : File table overflow
/export/home/prod/smc_smc/scripts/Monitor/CheckOutboundSmcInts.ss[78]: cannot make pipe


*************************************************
Cron: The previous message is the standard output
and standard error of one of your crontab commands:

/export/home/prod/smc_smc/scripts/Monitor/CheckOutboundSmcInts.ss

[A. Clay Stephenson]

It appears that you are hitting the kernel limit nfile. This may be a result of too many processes or processes that didn't terminate or it may indicate heavy usage. Use sar -v 5 5 to see how near the file-sz limit you are.

[OldSchool]

"cannot create", "cannot make pipe", "could not open file", "file table overflow"

man errno would seem to indicate that you have reached the max number of open files configured in the kernel on the system in question.

options are probably shutdown some processes or update kernel parms to allow more open files

[Patrick Wallek]

The key error, I believe, is "File table overflow". This means that you have reached the maximum number of open files on your system.

You will need to check your 'nfile' kernel parameter usage.

Do a:

# sar -v 5 5

And note the values in the 'file-sz' column. The value to the left of the '/' is the current usage. The value to the right is the limit. If the values are very close then you are probably occasionally hitting the limit.

To fix this you would need to increase the value of nfile, regenerate the kernel and reboot the system (11.11 and prior).