Simpler Navigation for Servers and Operating Systems
Completed: a much simpler Servers and Operating Systems section of the Community. We combined many of the older boards, so you won't have to click through so many levels to get at the information you need. Check the consolidated boards here as many sub-forums are now single boards.
Operating System - Tru64 Unix
cancel
Showing results for 
Search instead for 
Did you mean: 

weird ADVFS problem

BG Jeong
Advisor

weird ADVFS problem

Hi.
I am running ES40 2 node cluster with EMA12000 storage system.
there are 10 domains and 1 fileset per domain.

data1_domain#data1 /data1
data2_domain#data1 /data2
data3_domain#data1 /data3
........
data9_domain#data1 /data9
data10_domain#data1 /data10

there are about 3000 files per filesystem.
we are clearing files daily.
but,
all filesystem except /data3 are normal.
only /data3 filesystem is weird.

when i command 'rm *', it took five
hours to finish this command.

I checked filesystem using cfsmgr and drdmgr ....and messages file.
but there are no error, no warning.

Does anybody have any idea?
Tru64 from Korea
9 REPLIES
Ivan Ferreira
Honored Contributor

Re: weird ADVFS problem

Are you using some cloning feature for this fileset? Ensure that is not cloned when you run the command.
Por que hacerlo dificil si es posible hacerlo facil? - Why do it the hard way, when you can do it the easy way?
BG Jeong
Advisor

Re: weird ADVFS problem

thanks quickly reply.
there are not use cron fileset.
Tru64 from Korea
Michael Schulte zur Sur
Honored Contributor

Re: weird ADVFS problem

Hi,

is this a one time incident?
Have you checked the log in the HSG80?

greetings,

Michael
BG Jeong
Advisor

Re: weird ADVFS problem

Thacks.
HSG80 controller do not have any log.
Storage system ok.
??
Tru64 from Korea
Greg Yates
Valued Contributor

Re: weird ADVFS problem

Hi,

5 hours is a long time to delete 3000 files! :) Depending on what else may have been happening with that domain/controller, etc. I would tend to suspect the underlying storage or the domain. Try verify(8) or fixfdmn(8) with the "-n" option to check it. If it shows problems, use fixfdmn to fix it (or if you are completely cleaning out the domain anyway, go ahead and recreate it to eliminate the domain as a potential problem).

Even if the CFS server was a different member I wouldn't expect it to take this long... unless there were also cluster interconnect problems. But that should have caused other issues as well.

Greg
BG Jeong
Advisor

Re: weird ADVFS problem

Thanks your concerning!
I will try to verify or fixfdmn next monday
or remove and re-create domain.

is there any caution to use verify or fixfdmn?
Tru64 from Korea
Greg Yates
Valued Contributor

Re: weird ADVFS problem

I would only use verify to check the domain. If it needs repairing, fixfdmn is your best tool to fix it in place. Another tool, salvage, can be used to pull data from a severely damaged domain and recover it into a new domain but that's usually a last resort before the ultimate tool "vrestore" or similar. :)

Thanks,
Greg
Harmanjit_1
Frequent Advisor

Re: weird ADVFS problem

1) Please try to balance the domain if there are multiple disks are included.

2) Do defrag of the domain

These two will surely increase the performance and will help in your activity.

thanks
Michael Schulte zur Sur
Honored Contributor

Re: weird ADVFS problem

Hi,

do you mean there are no entries in the log? The HSG80 has a log of its own.

greetings,

Michael