Jacques,
The AIX command to change the Receive Pool Buffer size is:
chdev -l ent"x" -a rxbuf_pool_size=2048
where "x" is replace by the appropriate adapter number. However, it is not
as easy as just issuing this command. The adapter must be "offline" in order
for the change to take effect. This is sometimes a difficult state to
achieve. A possible circumvention, if you can arrange for a system recycle,
is to issue the command as:
chdev -l ent"x" -a rxbuf_pool_size=2048 -P
which says to update the ODM only. This can be done while the adapter is
active. You then perform your pre-arranged recycle of the system, and, when
it comes back up, the Receive Pool Buffer size will be set to the new value.
If that is not an option, then you will have to quiesce all traffic through
the adapter (don't forget sysinfod, if you happen to use that facility),
take the adapter down, issue the first listed command, and then put
everything back online. Clearly, I prefer the reboot option - far less
headaches.
Regards,
Jack Coyle
Rex Healthcare
(919) 784-3792
jack.coyle AT rexhealth DOT com
> ----------
> From: Jacques Butcher[SMTP:butcherjw AT ANANZI.CO DOT ZA]
> Reply To: ADSM: Dist Stor Manager
> Sent: Friday, August 15, 2003 2:01 AM
> To: ADSM-L AT VM.MARIST DOT EDU
> Subject: Re: Réf. : Re: Resolved: URGENT!! TSM client schedules
> notrunning (Guillaume Gilbert)
>
> Hi Guillaume,
>
> What command did you use to change the size of the "Receive
> Pool Buffer Size"?
>
> On Thu, 14 Aug 2003 12:44:09 -0400
> David E Ehresman <deehre01 AT LOUISVILLE DOT EDU> wrote:
> > If other people who are experiencing the problem also
> > have receive pool
> > buffer errors and increasing the size of the buffer pool
> > helps, maybe we
> > can get this added to the faq.
> >
> > David
> >
> > >>> guillaume.gilbert AT DESJARDINS DOT COM 08/14/03 11:50AM >>>
> > Hi David
> >
> > Thanks for the response. The "No receive pool buffer
> > errors" was at
> > 88934!!! I changed the Receive Pool Buffer Size to 2048
> > (it was at 384)
> > and am hoping everything will
> > be well for tonight. I am having the sysadmins bounce th
> > eschedulers on
> > all the servers that did not work last night just in
> > case...
> >
> >
> > Guillaume Gilbert
> > Conseiller - Gestion du stockage
> > CGI - Gestion intigri des technologies
> > TEL.: (514) 415-3000 ext 5091
> > Pager: (514) 957-2615
> > Email : guillaume.gilbert AT cgi DOT com
> >
> ____________________________________________________________
> >
> >
> >
> >
> > David E Ehresman <deehre01 AT LOUISVILLE DOT EDU>@VM.MARIST.EDU>
> > on 2003-08-14
> > 11:04:49
> >
> > Veuillez ripondre ` "ADSM: Dist Stor Manager"
> > <ADSM-L AT VM.MARIST DOT EDU>
> >
> > Envoyi par : "ADSM: Dist Stor Manager"
> > <ADSM-L AT VM.MARIST DOT EDU>
> >
> >
> > Pour : ADSM-L AT VM.MARIST DOT EDU
> > cc :
> > Objet : Re: Resolved: URGENT!! TSM client schedules
> > not running
> >
> > My problem was also with Prompted nodes only. Polling
> > nodes ran ok as
> > did croned backups.
> >
> > The tcp setting was the size of the Receive Pool Buffer.
> > IBM gave me
> > this query to determine if there was a problem:
> > (echo "----";date; netstat -v ent1 | grep "Receive Pool
> > Buffer" )
> >
> > The "No Receive Pool Buffer Errors" line should report
> > zero. They had
> > me set the "Receive Pool Buffer Size" up to 2048 (AIX
> > 5.1).
> > Unfortunately, I don't remember the command to do that.
> >
> > David
> >
> > >>> guillaume.gilbert AT DESJARDINS DOT COM 08/14/03 10:29AM >>>
> > Ok you guys had to start talking about this and now it
> > has happened to
> > me. Nearly all my servers in prompt mode didn't take a
> > backup last
> > night. Nothing in the activity
> > log and nothing in the dsmsched.logs. I'm running server
> > 5.1.6.4 on
> > AIX
> > 5.1 and clients range from 4.1.2.12 to 5.1.5.15, all
> > Windows (NT and
> > 2k). Most of these servers
> > had the MS patch installed for the blaster virus
> > yesterday. Is this
> > just a coincidence??. I know its not a network problem
> > because all my
> > UNIX clients where running
> > backups all night and those are not scheduled by TSM but
> > by crontabs.
> >
> > So David, what was the tcpip settings you had to change?
> >
> > Thanks
> >
> > Guillaume Gilbert
> > Conseiller - Gestion du stockage
> > CGI - Gestion intigri des technologies
> > TEL.: (514) 415-3000 ext 5091
> > Pager: (514) 957-2615
> > Email : guillaume.gilbert AT cgi DOT com
> >
> ____________________________________________________________
> >
> > When I had the problem, I did open a PMR and they finally
> > tracked my
> > problem down to a tcp read buffer problem. They had me
> > adjust the tcp
> > values and the problem has not reoccurred.
> >
> > David
>
> Jacques Butcher
> TCM (Technology Corporate Management) Software Engineer
> Cell: +27 (0)84 676 0329
> Tell: +27 (0)11 483-2000
> Fax: +27 (0)11 728-3656
> Nat. IT Diploma, MCSE, IBM Tivoli Storage Manager 5.1
> Certified, NetVault Certified, IPSTor Certified,
> IBM Certified Specialist - Enterprise Tape Solutions
> Version 2
>
> ==
> Download ringtones, logos and picture messages at Ananzi Mobile Fun.
> http://www.ananzi.co.za/cgi-bin/goto.pl?mobile
>
|
