RE: NETTYPE in IDS 10
Posted in 2011
The poster asked whether the user count in NETTYPE (e.g. soctcp,5,300,NET) is a hard limit in IDS 10 or whether the server will accept more connections dynamically. Answer: shared memory (ipcshm) connections are static, while TCP and other network protocols expand dynamically. The NETTYPE user number mainly sizes the pre-allocated network buffers; beyond it, buffers are allocated and freed per message, costing performance, with the real cap being system resources. Guidance given: roughly 300-350 connections per poll thread before performance degrades.
Auto-generated by DrWatson from the posts below — may be imperfect; read the full thread.
Topics: General Discussion
Thank you. What is the IDS 10 rule now ? Is TCP dynamic ?
Dirk
> -----Original Message-----
> From: ids-bounces@iiug.org [mailto:ids-bounces@iiug.org] On Behalf Of
> Obnoxio The Clown
> Sent: Monday, 13 June 2011 01:16 PM
> To: ids@iiug.org
> Subject: Re: NETTYPE in IDS 10 [23999]
>
> On 13/06/2011 11:17, Dirk Cornel.... wrote:
> > NETTYPE ipcshm,2,300,CPU # Configure poll thread(s) for nettype>
> This one won't.
>
> > NETTYPE soctcp,5,300,NET # Configure poll thread(s) for nettype>
> This one will.
>
> > Question: Is the tcp NETTYPE in IDS 10 dynamic, ie. Will the database
> grab
> > more sessions if it needs more, even though your NETTYPE is set to a
> fixed
> > value ?
>
> --
> Cheers,
> Obnoxio The Clown
>
> http://obotheclown.blogspot.com
> I will now proceed to pleasure myself with this fish.
>
>
> ***********************************************************************
> ********
> Forum Note: Use "Reply" to post a response in the discussion forum.
NOTE: This e-mail message is subject to the MTN Group disclaimer see
http://www.mtn.co.za/SUPPORT/LEGAL/Pages/EmailDisclaimer.aspx
On 13/06/2011 13:00, Dirk Cornel.... wrote: > Thank you. What is the IDS 10 rule now ? Is TCP dynamic ? It always has been. SHM has always been static, TCP has always been dynamic. -- Cheers, Obnoxio The Clown http://obotheclown.blogspot.com I will now proceed to pleasure myself with this fish.
Thanks. I am very rusty at the moment. But I'll catch up again. > -----Original Message----- > From: ids-bounces@iiug.org [mailto:ids-bounces@iiug.org] On Behalf Of > Obnoxio The Clown > Sent: Monday, 13 June 2011 04:31 PM > To: ids@iiug.org > Subject: Re: NETTYPE in IDS 10 [24004] > > On 13/06/2011 13:00, Dirk Cornel.... wrote: > > Thank you. What is the IDS 10 rule now ? Is TCP dynamic ? > > It always has been. SHM has always been static, TCP has always been > dynamic. > > -- > Cheers, > Obnoxio The Clown > > http://obotheclown.blogspot.com > I will now proceed to pleasure myself with this fish. > > > *********************************************************************** > ******** > Forum Note: Use "Reply" to post a response in the discussion forum. NOTE: This e-mail message is subject to the MTN Group disclaimer see http://www.mtn.co.za/SUPPORT/LEGAL/Pages/EmailDisclaimer.aspx
One last question: If it is dynamic, what will the limit then be ? ...... the max number of connections / processes in the Unix kernel ? > -----Original Message----- > From: ids-bounces@iiug.org [mailto:ids-bounces@iiug.org] On Behalf Of > Obnoxio The Clown > Sent: Monday, 13 June 2011 04:31 PM > To: ids@iiug.org > Subject: Re: NETTYPE in IDS 10 [24004] > > On 13/06/2011 13:00, Dirk Cornel.... wrote: > > Thank you. What is the IDS 10 rule now ? Is TCP dynamic ? > > It always has been. SHM has always been static, TCP has always been > dynamic. > > -- > Cheers, > Obnoxio The Clown > > http://obotheclown.blogspot.com > I will now proceed to pleasure myself with this fish. > > > *********************************************************************** > ******** > Forum Note: Use "Reply" to post a response in the discussion forum. NOTE: This e-mail message is subject to the MTN Group disclaimer see http://www.mtn.co.za/SUPPORT/LEGAL/Pages/EmailDisclaimer.aspx
The number of users specified in the NETTYPE configuration parameter for TCP connections is there to optimize the number of network buffers which will be allocated. If you exceed the number then each network buffer will be allocated and free'd on each message. The penalty being performance and the limit is really determined by resources availability. John F. Miller III STSM, Embedability Architect miller3@us.ibm.com 503-578-5645 IBM Informix Dynamic Server (IDS) ids-bounces@iiug.org wrote on 06/13/2011 08:31:45 AM: > [image removed] > > RE: NETTYPE in IDS 10 [24007] > > Dirk Cornel.... > > to: > > ids > > 06/13/2011 08:32 AM > > Sent by: > > ids-bounces@iiug.org > > Please respond to ids > > One last question: If it is dynamic, what will the limit then be ? ...... the > max number of connections / processes in the Unix kernel ? > > > -----Original Message----- > > From: ids-bounces@iiug.org [mailto:ids-bounces@iiug.org] On Behalf Of > > Obnoxio The Clown > > Sent: Monday, 13 June 2011 04:31 PM > > To: ids@iiug.org > > Subject: Re: NETTYPE in IDS 10 [24004] > > > > On 13/06/2011 13:00, Dirk Cornel.... wrote: > > > Thank you. What is the IDS 10 rule now ? Is TCP dynamic ? > > > > It always has been. SHM has always been static, TCP has always been > > dynamic. > > > > -- > > Cheers, > > Obnoxio The Clown > > > > http://obotheclown.blogspot.com > > I will now proceed to pleasure myself with this fish. > > > > > > *********************************************************************** > > ******** > > Forum Note: Use "Reply" to post a response in the discussion forum. > > NOTE: This e-mail message is subject to the MTN Group disclaimer see > http://www.mtn.co.za/SUPPORT/LEGAL/Pages/EmailDisclaimer.aspx > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. >
Shared memory connections do not expand dynamically, all other protocols do.
Art
Art S. Kagel
Advanced DataTools (www.advancedatatools.com)
Blog: http://informix-myview.blogspot.com/
Disclaimer: Please keep in mind that my own opinions are my own opinions and
do not reflect on my employer, Advanced DataTools, the IIUG, nor any other
organization with which I am associated either explicitly, implicitly, or by
inference. Neither do those opinions reflect those of other individuals
affiliated with any entity with which I am affiliated nor those of the
entities themselves.
On Mon, Jun 13, 2011 at 8:00 AM, Dirk Cornel.... <moolma_dc@mtn.co.za>wrote:
> Thank you. What is the IDS 10 rule now ? Is TCP dynamic ?
>
> Dirk
>
> > -----Original Message-----
> > From: ids-bounces@iiug.org [mailto:ids-bounces@iiug.org] On Behalf Of
> > Obnoxio The Clown
> > Sent: Monday, 13 June 2011 01:16 PM
> > To: ids@iiug.org
> > Subject: Re: NETTYPE in IDS 10 [23999]
> >
> > On 13/06/2011 11:17, Dirk Cornel.... wrote:
> > > NETTYPE ipcshm,2,300,CPU # Configure poll thread(s) for nettype> >
> > This one won't.
> >
> > > NETTYPE soctcp,5,300,NET # Configure poll thread(s) for nettype> >
> > This one will.
> >
> > > Question: Is the tcp NETTYPE in IDS 10 dynamic, ie. Will the database
> > grab
> > > more sessions if it needs more, even though your NETTYPE is set to a
> > fixed
> > > value ?
> >
> > --
> > Cheers,
> > Obnoxio The Clown
> >
> > http://obotheclown.blogspot.com
> > I will now proceed to pleasure myself with this fish.
> >
> >
> > ***********************************************************************
> > ********
> > Forum Note: Use "Reply" to post a response in the discussion forum.
>
> NOTE: This e-mail message is subject to the MTN Group disclaimer see
> http://www.mtn.co.za/SUPPORT/LEGAL/Pages/EmailDisclaimer.aspx
>
>
>
>
*******************************************************************************
> Forum Note: Use "Reply" to post a response in the discussion forum.
>
>
--bcaec501634995719404a59cf2e7
No real max IB, however, above about 350 connections per poll thread connection performance starts to degrade. IBM recommends 300-350 connections per thread. Art Art S. Kagel Advanced DataTools (www.advancedatatools.com) Blog: http://informix-myview.blogspot.com/ Disclaimer: Please keep in mind that my own opinions are my own opinions and do not reflect on my employer, Advanced DataTools, the IIUG, nor any other organization with which I am associated either explicitly, implicitly, or by inference. Neither do those opinions reflect those of other individuals affiliated with any entity with which I am affiliated nor those of the entities themselves. On Mon, Jun 13, 2011 at 11:31 AM, Dirk Cornel.... <moolma_dc@mtn.co.za>wrote: > One last question: If it is dynamic, what will the limit then be ? ...... > the > max number of connections / processes in the Unix kernel ? > > > -----Original Message----- > > From: ids-bounces@iiug.org [mailto:ids-bounces@iiug.org] On Behalf Of > > Obnoxio The Clown > > Sent: Monday, 13 June 2011 04:31 PM > > To: ids@iiug.org > > Subject: Re: NETTYPE in IDS 10 [24004] > > > > On 13/06/2011 13:00, Dirk Cornel.... wrote: > > > Thank you. What is the IDS 10 rule now ? Is TCP dynamic ? > > > > It always has been. SHM has always been static, TCP has always been > > dynamic. > > > > -- > > Cheers, > > Obnoxio The Clown > > > > http://obotheclown.blogspot.com > > I will now proceed to pleasure myself with this fish. > > > > > > *********************************************************************** > > ******** > > Forum Note: Use "Reply" to post a response in the discussion forum. > > NOTE: This e-mail message is subject to the MTN Group disclaimer see > http://www.mtn.co.za/SUPPORT/LEGAL/Pages/EmailDisclaimer.aspx > > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. > > --20cf3071c696a8d4b004a59d49a1