Well, not really any answers to this yet. I got two responses from
people who have experienced this same problem.
esh@ibx.com (Bill Eshbach)
z85030@uprc.com (Slezak L. P. (Paul))
Their workaround is not to use the diskinfo agents to do any
monitoring. Paul Slezak said he had the same problem on SNM 2.1
monitoring a 4/670 running 4.1.3. I suspect that it uses the same
na.diskinfo agent as SNM 2.0 for monitoring SunOS 4.x systems.
What I have found is a bug in SunSolve:
Bug Id: 1110613
Category: net_manager
Subcategory: agent
Release summary: 2.0
Synopsis: many na.diskinfo processes running on client machine
This bug was supposed to be fixed in the SNM patch 100770-02 (now at
revision -06). Obviously it was not completely fixed. We have a service
call in with Sun, so we'll have to wait and see what the engineers at
SunConnect have to say.
regards,
--
Glenn Satchell glenn@uniq.com.au | "This is a unix system.
Uniq Professional Services Pty Ltd ACN 056 279 335 | I can do this easy."
PO Box 70, Paddington, NSW 2021, (Sydney) Australia |
Phone 02 360 7434 Pager 016 287 000 Fax 02 331 2572 | - Lex, Jurassic Park
"Sun Accredited System Consultants" |
>
> Hi folks,
>
> We are running SunNet Manager 2.0 on an SS2 with SunOS 4.1.3.
>
> I have a strange problem where client machines are hanging or crashing
> due to many na.diskinfo processes running on them. In one instance
> there was over 280 processes running on a SS690 that normally sits
> around the 100 mark. SNM patch 100770-06 has been installed on all the
> machines without making any noticeable difference.
>
> The problem occurs on all the machines where we run na.diskinfo. We
> have set up an event to check for the disk capacity exceeding 85% and a
> second event at higher priority when it exceeds 95%. The client
> machines are SS2, IPX, SS10, SS690, all running 4.1.3.
>
> The na.diskinfo will behave normally for upto several hours or days,
> then in a period as short as 5 minutes it will misbehave and there will
> be 40, 50 or more copies of it running.
>
> There are no error messages in /var/adm/messages until the process
> table fills up or we run out of swap.
>
> There aren't a lot of patches installed on these systems. The sun4m's
> have patch 100726 installed (sorry don't know which revision but I
> don't think it is a problem since the sun4c's behave the same way), SNM
> patch 100770-02 was installed initially with SNM, then upgraded to the
> -06 version, but otherwise they are pretty clean.
>
> Any ideas, suggestions, or further questions are welcome. I'll
> summarise results to the list.
>
> regards,
> --
> Glenn Satchell glenn@uniq.com.au | "This is a unix system.
> Uniq Professional Services Pty Ltd ACN 056 279 335 | I can do this easy."
> PO Box 70, Paddington, NSW 2021, (Sydney) Australia |
> Phone 02 360 7434 Pager 016 287 000 Fax 02 331 2572 | - Lex, Jurassic Park
>
This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:08:57 CDT