UNSOLVED

razvan2

updated

21 years ago

R

razvan2

2 Intern

•

181 Posts

0

622

March 7th, 2006 04:00

NW 7.3 problems

This morning I notice that many backup failed and the server has a very high load.
I try ps to see the process and I notice that are 5 savegrp in the memory, and mminfo is stuck, dosen¿t display nothing.
After I issue nsr_shutdown ¿q the load remain high, and I notice that my 5 savegrp are still running.
Also the load remain very high.
After I kill savegrp process the load became normal.

I encounter this strange situation after I migrate the 7.3.
  • ble1

    6 Operator

    •

    14354 Posts

    •

    56186 Points

    168

    0

    Posted March 7th, 2006 04:00

    This might not be related to 7.3 at all. I wonder at what stage this savegrps were stuck (media db or index db update). Do you see anything ususual in system logs? Next time, if that happens again, try to run truss against the nsrd to see if which state it is.
  • razvan2

    2 Intern

    •

    181 Posts

    168

    0

    Posted March 7th, 2006 05:00

    I notice that some of my groups start with 1h delay (instead of 5.10 AM start at 6.25 AM)
    Also in initial state, I try to issue mminfo without any result. Maybe I have problem with media database.
    About your advice to run nsrd, could you give me more details?
    Thx
    Razvan
  • ble1

    6 Operator

    •

    14354 Posts

    •

    56186 Points

    168

    0

    Posted March 7th, 2006 05:00

    I assume this is UNIX... so run truss -pf $NSRD_PID to verify if nsrd is doing any activity at all during this strange state.
  • ble1

    6 Operator

    •

    14354 Posts

    •

    56186 Points

    168

    0

    Posted March 7th, 2006 06:00

    Strace will do :)
  • razvan2

    2 Intern

    •

    181 Posts

    168

    0

    Posted March 7th, 2006 06:00

    Is linux, i will use strace -p.
  • razvan2

    2 Intern

    •

    181 Posts

    168

    0

    Posted March 7th, 2006 06:00

    My OS is linux, I will use strace.
  • razvan2

    2 Intern

    •

    181 Posts

    168

    0

    Posted March 7th, 2006 07:00

    I have more information
    Look what I see on a client daemon.log:
    03/07/06 06:32:04 nsrexecd: RPC error: Can't setup SSL session channel to host: legatosrv for GSS Legato authentication.: Timed out
  • ble1

    6 Operator

    •

    14354 Posts

    •

    56186 Points

    168

    0

    Posted March 7th, 2006 09:00

    If nsrd for some reason was blocked then that would be one consequence too I would assume.