[prev in list] [next in list] [prev in thread] [next in thread] 

List:       net-snmp-support
Subject:    [ net-snmp-Support Requests-422389 ] 25% cpu usage (100 on one of 4 cpu's)
From:       noreply () sourceforge ! net
Date:       2001-05-14 17:37:01
[Download RAW message or body]

Support Requests item #422389, was updated on 2001-05-08 10:10
You can respond by visiting: 
http://sourceforge.net/tracker/?func=detail&atid=212694&aid=422389&group_id=12694

Category: None
Group: None
Status: Open
Priority: 5
Submitted By: Bill Schwanitz (bilsch)
Assigned to: John Naylon (jbpn)
Summary: 25% cpu usage (100 on one of 4 cpu's)

Initial Comment:
I have noticed on one of my boxes the snmpd proc. is
taking 100% of one of 4 cpu's.  It's a Sun Ent. 4500, 4
cpu box.

Is there any way to get the snmpd proc to tell me what
the heck it's doing besides trussing the app? I've run
this to attempt to: snmpd -fVLD

Thanks,

Bill Schwanitz
AT&T Solutions

----------------------------------------------------------------------

>Comment By: Bill Schwanitz (bilsch)
Date: 2001-05-14 10:37

Message:
Logged In: YES 
user_id=214814

I have tried to get onto the CVS tree to snag the current
snapshot, I may be doing something wrong, I may not actually
be able to get out my firewall.. Is there an easier/other
way to obtain the source??

----------------------------------------------------------------------

Comment By: Bill Schwanitz (bilsch)
Date: 2001-05-14 09:46

Message:
Logged In: YES 
user_id=214814

Yea, it's an E4500, 4 cpu, 1.2 gig ram, solaris 2.7

here is output from truss.. it repeats over and over.

getmsg(10, 0x00000000, 0xFFBE1EB0, 0xFFBE1EAC)  = 0
getmsg(10, 0xFFBE1EB0, 0x00000000, 0xFFBE1EAC)  = 2
getmsg(10, 0x00000000, 0xFFBE1EB0, 0xFFBE1EAC)  = 0
getmsg(10, 0xFFBE1EB0, 0x00000000, 0xFFBE1EAC)  = 2
getmsg(10, 0x00000000, 0xFFBE1EB0, 0xFFBE1EAC)  = 0
getmsg(10, 0xFFBE1EB0, 0x00000000, 0xFFBE1EAC)  = 2
getmsg(10, 0x00000000, 0xFFBE1EB0, 0xFFBE1EAC)  = 0
getmsg(10, 0xFFBE1EB0, 0x00000000, 0xFFBE1EAC)  = 2
getmsg(10, 0x00000000, 0xFFBE1EB0, 0xFFBE1EAC)  = 0
getmsg(10, 0xFFBE1EB0, 0x00000000, 0xFFBE1EAC)  = 2
getmsg(10, 0x00000000, 0xFFBE1EB0, 0xFFBE1EAC)  = 0
ioctl(10, I_FLUSH, FLUSHRW)                     = 0
time()                                          = 989858760
ioctl(9, 0xC0206911, 0x000DDA10)                = 0
ioctl(9, 0xC0206916, 0x000DDA10)                = 0
ioctl(5, KSTAT_IOC_READ, "kstat_headers")       = 222
ioctl(9, 0xC020690D, 0x000DDA10)                = 0
time()                                          = 989858760
putmsg(10, 0xFFBE1EB0, 0x00000000, 0)           = 0
getmsg(10, 0xFFBE1EB0, 0x00000000, 0xFFBE1EAC)  = 2
getmsg(10, 0x00000000, 0xFFBE1EB0, 0xFFBE1EAC)  = 0
getmsg(10, 0xFFBE1EB0, 0x00000000, 0xFFBE1EAC)  = 2
getmsg(10, 0x00000000, 0xFFBE1EB0, 0xFFBE1EAC)  = 0
getmsg(10, 0xFFBE1EB0, 0x00000000, 0xFFBE1EAC)  = 2
getmsg(10, 0x00000000, 0xFFBE1EB0, 0xFFBE1EAC)  = 0
getmsg(10, 0xFFBE1EB0, 0x00000000, 0xFFBE1EAC)  = 2
getmsg(10, 0x00000000, 0xFFBE1EB0, 0xFFBE1EAC)  = 0
getmsg(10, 0xFFBE1EB0, 0x00000000, 0xFFBE1EAC)  = 2
getmsg(10, 0x00000000, 0xFFBE1EB0, 0xFFBE1EAC)  = 0
getmsg(10, 0xFFBE1EB0, 0x00000000, 0xFFBE1EAC)  = 2
getmsg(10, 0x00000000, 0xFFBE1EB0, 0xFFBE1EAC)  = 0
getmsg(10, 0xFFBE1EB0, 0x00000000, 0xFFBE1EAC)  = 2
getmsg(10, 0x00000000, 0xFFBE1EB0, 0xFFBE1EAC)  = 0
getmsg(10, 0xFFBE1EB0, 0x00000000, 0xFFBE1EAC)  = 2
getmsg(10, 0x00000000, 0xFFBE1EB0, 0xFFBE1EAC)  = 0
getmsg(10, 0xFFBE1EB0, 0x00000000, 0xFFBE1EAC)  = 2
getmsg(10, 0x00000000, 0xFFBE1EB0, 0xFFBE1EAC)  = 0
getmsg(10, 0xFFBE1EB0, 0x00000000, 0xFFBE1EAC)  = 2
getmsg(10, 0x00000000, 0xFFBE1EB0, 0xFFBE1EAC)  = 0
getmsg(10, 0xFFBE1EB0, 0x00000000, 0xFFBE1EAC)  = 2
getmsg(10, 0x00000000, 0xFFBE1EB0, 0xFFBE1EAC)  = 0
getmsg(10, 0xFFBE1EB0, 0x00000000, 0xFFBE1EAC)  = 2
getmsg(10, 0x00000000, 0xFFBE1EB0, 0xFFBE1EAC)  = 0
getmsg(10, 0xFFBE1EB0, 0x00000000, 0xFFBE1EAC)  = 2
getmsg(10, 0x00000000, 0xFFBE1EB0, 0xFFBE1EAC)  = 0
getmsg(10, 0xFFBE1EB0, 0x00000000, 0xFFBE1EAC)  = 2
getmsg(10, 0x00000000, 0xFFBE1EB0, 0xFFBE1EAC)  = 0
ioctl(10, I_FLUSH, FLUSHRW)                     = 0
time()                                          = 989858760
ioctl(9, 0xC0206911, 0x000DDA30)                = 0
ioctl(9, 0xC0206916, 0x000DDA30)                = 0
ioctl(5, KSTAT_IOC_READ, "kstat_headers")       = 222
ioctl(9, 0xC020690D, 0x000DDA30)                = 0
^Ctime()                                                =
989858760


----------------------------------------------------------------------

Comment By: John Naylon (jbpn)
Date: 2001-05-14 07:00

Message:
Logged In: YES 
user_id=93926

As fun as that sounds, I don't think I would be able to tell
if the problem is where I suggested from a core dump.

You could truss it and see if it's calling select(); recv()
repeatedly when it hits 100%.  Do you have a repeatable
method for getting it into the 100% CPU state, or does it
occur more or less at random?  If so, it would be
particularly interesting to try the CVS version.  I assume
your Sun box is running Solaris?  Which version?

----------------------------------------------------------------------

Comment By: Bill Schwanitz (bilsch)
Date: 2001-05-14 06:54

Message:
Logged In: YES 
user_id=214814

Yea, the SNMP daemon still responds.  It's a tad bit slow,
but that is probably because it is going nuts. 

Did you want to take a look at a core image? I can grab the
core from the process and send it to you.

----------------------------------------------------------------------

Comment By: John Naylon (jbpn)
Date: 2001-05-14 06:26

Message:
Logged In: YES 
user_id=93926

You could try the latest CVS source.  There was an issue
whereby the main select() loop could get into a tight loop,
and I don't think the fix for this made it into 4.2.1.  Of
course this may not be the cause of the problem, but it's
worth a try.  Is the snmpd still responsive to requests?

----------------------------------------------------------------------

Comment By: Bill Schwanitz (bilsch)
Date: 2001-05-14 06:14

Message:
Logged In: YES 
user_id=214814

Version is both 4.2 and 4.2.1

----------------------------------------------------------------------

Comment By: John Naylon (jbpn)
Date: 2001-05-14 02:14

Message:
Logged In: YES 
user_id=93926

What version of snmpd is this?

----------------------------------------------------------------------

You can respond by visiting: 
http://sourceforge.net/tracker/?func=detail&atid=212694&aid=422389&group_id=12694

_______________________________________________
Net-snmp-support mailing list
Net-snmp-support@lists.sourceforge.net
http://lists.sourceforge.net/lists/listinfo/net-snmp-support

[prev in list] [next in list] [prev in thread] [next in thread] 

Configure | About | News | Add a list | Sponsored by KoreLogic