[Asterisk-Users] Dual T400P, SMP, performance issues

The Traveller traveler at xs4all.nl
Mon Jun 16 12:45:54 MST 2003


Yo,

I've seen very similar Zaptel-related freezes on a wide variety of
mainboards (SMP as well as non-SMP), with X100P's as well as with an E100P.
At some point, almost always at the moment a call through one of those cards
connects or disconnects, the machine completely stops responding and needs
a reset to come back to life.  A very nice way to trigger it with the E100P
seems to be to put around 10-20 channels of it into a meetme-conference and
then issue the "stop now"-command on the Asterisk-console.  A high volume
of connects / disconnects seems to trigger the freezes.  I'm still
investigating the issue and am going to try different kernels and
some custom kernel-patches.

One of my boxes (dual PIII-750, Intel L440GX+-board) with an X100P and
a TDM40P in it hasn't frozen since I installed kernel 2.4.21-rc2 with
the ACPI-patch (http://sourceforge.net/projects/acpi/).  I'll probably try
that on the box with the E100P first.  Be sure enable "Power Management
support" in your kernel-config, disable APM, enable ACPI and check all
ACPI-options, except for "CPU Enumeration Only".  Note that this ACPI-
patch also handles IRQ-routing and might help in cases where the BIOS assigns
the same IRQ to some devices (or, as was the case for me, none at all).



    Grtz,

      Oliver

On Mon, Jun 16, 2003 at 13:03:20 -0500, Alex Zarubin wrote:

> Mark,
> 
> As far as pings - we have cases when we could ping the box on both
> interfaces and there are cases when we could not (we tried 3-4 sets of
> NICs and drivers). All telnets, X, ssh etc. are definitely dead.
> No coredumps (asterisk was started with -g option), no kernel panics.
> Black console, Alt-SysRq combinations don't work.
> Pretty much no options but rebooting the box.
> 
> As far as SMP and single T400P - we'll try and report the results
> but the idea was to go with as high density as possible ...
> 
> What do you think of using hyperthreading - should we enable or disable it
> for the box running asterisk?
> 
> What about -DCONFIG_ZAPTEL_WATCHDOG ? Can it help and how to use it?
> 
> Thank you.
> Alex Zarubin
> 
> -----Original Message-----
> From: Mark Spencer [mailto:markster at digium.com]
> Sent: Saturday, June 14, 2003 10:23 AM
> To: 'asterisk-users at lists.digium.com'
> Subject: RE: [Asterisk-Users] Dual T400P, SMP, performance issues
> 
> 
> When you say "stops responding" do you mean no more pings, telnet dead,
> etc?  Or do you mean asterisk stops responding?  Is there a segfault or
> kernel panic, or any other failure diagnostic?
> 
> Mark
> 
> On Thu, 12 Jun 2003, Alex Zarubin wrote:
> 
> > Zaptel was compiled with -D__SMP__
> >
> > We've installed irqbalance and the picture improved a lot
> > (thanks to Jared Smith). Do you still see problems in our
> /proc/interrupts?
> >
> > The big issue for us now is that after 24+ hours of the test load PRI->SIP
> > our Dell PE2650, dual 2.6 GHz Xeon, 2 Gb RAM, 2 T400P, 2.4.20-18.7smp #1
> SMP
> > stops responding to anything.
> >
> > So the questions are:
> > 	- are there known issues with PE2650 and ways to fix them?
> > 	- can someone recommend the 'stable' 2.4 SMP kernel for this
> > 	  kind of load?
> > 	- any expertise in this area will be appreciated
> >
> >            CPU0       CPU1       CPU2       CPU3
> >   0:     230710      30030      50050          0    IO-APIC-edge  timer
> >   1:          5          0          0        233    IO-APIC-edge  keyboard
> >   2:          0          0          0          0          XT-PIC  cascade
> >   5:          0          0          0          0   IO-APIC-level  usb-ohci
> >   8:          1          0          0          0    IO-APIC-edge  rtc
> >  14:         27          0          2          0    IO-APIC-edge  ide0
> >  20:    2085442     400221          0     230232   IO-APIC-level  tor2
> >  24:     293848    1841658      10010     570568   IO-APIC-level  tor2
> >  28:          5      25643          0          0   IO-APIC-level  eth0
> >  29:          5          0    5165040          0   IO-APIC-level  eth1
> >  30:      43720      35467       1291       3296   IO-APIC-level  aacraid
> > NMI:          0          0          0          0
> > LOC:     310618     310616     310616     310616
> > ERR:          0
> > MIS:          0
> >
> > Thank you.
> > Alex Zarubin
> >
> > -----Original Message-----
> > From: Martin Pycko [mailto:martinp at digium.com]
> > Sent: Tuesday, June 10, 2003 9:48 AM
> > To: 'asterisk-users at lists.digium.com'
> > Subject: Re: [Asterisk-Users] Dual T400P, SMP, performance issues
> >
> >
> > Are you sure that you compiled zaptel for __SMP__ ?
> > Edit your zaptel/Makefile.
> >
> >   0:   75283844   75241320   75286285   75247088    IO-APIC-edge  timer
> >   1:          1          0          1          1    IO-APIC-edge  keyboard
> >   2:          0          0          0          0          XT-PIC  cascade
> >   3:          0          0          0          0   IO-APIC-level  usb-ohci
> >   8:          1          0          0          0    IO-APIC-edge  rtc
> >  15:          1          0          0          1    IO-APIC-edge  ide1
> >  16:   22134870   22120997   22135905   22122829   IO-APIC-level  eth0
> >  25:       4670       4548       4614       4518   IO-APIC-level  tor2
> >
> > All the four CPU's should have IRQ's like in the example above.
> >
> > Martin
> >
> > On Mon, 9 Jun 2003, Alex Zarubin wrote:
> >
> > > Hi,
> > >
> > > We are trying to validate Asterisk as a media gateway PRI <-> SIP with
> two
> > > T400P (8 T1s) per box. The first
> > > experience with BOX1 (Compaq, 2.53 GHz, 1 Gb RAM) and just one T400P was
> > > encouraging - on the load
> > > test with 3 T1s worth of calls we had on average 75% idle CPU.
> > >
> > > Not so with BOX2 (Dell, single 2.6 GHz Xeon, 1 Gb RAM, 2 T400P) and BOX3
> > > (Dell, dual 2.6 GHz Xeon,
> > > 2 Gb RAM, 2 T400P, asterisk/zaptel is built with SMP support).
> > >
> > > On the similar load test (as with the BOX1) BOX2 was showing 0% idle CPU
> > 70%
> > > of the time. Just 3 T1s
> > > out of 8.
> > >
> > > On the load test with just 2 T1s BOX3 was very close to 0% idle on CPU0,
> > > CPU1 was at 95% idle.
> > > The process ksoftirqd_CPU0 was close to the top of the 'top', with
> > > /proc/interrupts showing tor2 related
> > > numbers growing very fast. We had 2 T1s plugged into the first T400P
> > board,
> > > with nothing going into the second,
> > > but the number of interrupts for the both boards was growing at the same
> > > pace. Here are the interrupts
> > > (after the box reboot, so they are not that big as they were) - do they
> > look
> > > OK?
> > >
> > >
> > >             CPU0       CPU1       CPU2       CPU3
> > >   0:     122556          0          0          0    IO-APIC-edge  timer
> > >   1:          4          0          0          0    IO-APIC-edge
> keyboard
> > >   2:          0          0          0          0          XT-PIC
> cascade
> > >   5:          0          0          0          0   IO-APIC-level
> usb-ohci
> > >   8:          1          0          0          0    IO-APIC-edge  rtc
> > >  12:         20          0          0          0    IO-APIC-edge  PS/2
> > Mouse
> > >  14:         23          0          2          0    IO-APIC-edge  ide0
> > >  20:     516930          0          0          0   IO-APIC-level  tor2
> > >  24:     516524          0          0          0   IO-APIC-level  tor2
> > >  28:      10600          0          0          0   IO-APIC-level  eth0
> > >  29:       4837          0          0          0   IO-APIC-level  eth1
> > >  30:      24831          0          0          0   IO-APIC-level
> aacraid
> > > NMI:          0          0          0          0
> > > LOC:     122430     122429     122429     122428
> > > ERR:          0
> > > MIS:          0
> > >
> > > Not sure what went wrong. Any suggestions on how to work with 2 T400P in
> a
> > > box (without hurting performance)
> > > and how to get advantage of SMP for Asterisk would be appreciated.
> > >
> > > Any known Linux kernel related issues (2.4.20-13.7smp #1 SMP for BOX3 )?
> > >
> > > Thank you.
> > >
> > > Alex Zarubin
> > >
> > >
> > >
> >
> > _______________________________________________
> > Asterisk-Users mailing list
> > Asterisk-Users at lists.digium.com
> > http://lists.digium.com/mailman/listinfo/asterisk-users
> >
> 
> _______________________________________________
> Asterisk-Users mailing list
> Asterisk-Users at lists.digium.com
> http://lists.digium.com/mailman/listinfo/asterisk-users



More information about the asterisk-users mailing list