[Asterisk-Users] zttest
Mike Mueller
mmueller at ss7box.com
Wed May 18 20:39:43 MST 2005
On Mon, May 16, 2005 at 08:59:23PM +1200, Damian Funnell wrote:
> Hi Waldo, I would be money on your problem being related to the accuracy
> of zttest. One way of checking IRQ's is to run "cat /proc/interrupts",
> but it is a lot more accurate to run "lspci -v" and "lspci -vb".
>
> I would recommend Googling the lspci command, although the output is
> pretty self explanatory. The TDM appears as a TigerJet card, not sure
> what TE410P will list as.
>
> PCI devices have their IRQ's dictated by the BIOS of the host system.
> How (and if) you can configure these manually depends on the type of
> BIOS you have... in our IBM xSeries 206 we had to actually juggle cards
> between slots to get it to assign a unique IRQ to the TDM400P.
>
> Good luck!
Before changing IRQ in BIOS I had:
4: 23643 XT-PIC eth0, wctdm
After changing IRQ in BIOS I had:
b2:/usr/src/zaptel# cat /proc/interrupts
CPU0
0: 51108 XT-PIC timer
1: 2 XT-PIC keyboard
2: 0 XT-PIC cascade
7: 23643 XT-PIC eth0, wctdm
14: 3073 XT-PIC ide0
15: 2 XT-PIC ide1
NMI: 0
LOC: 51070
ERR: 0
MIS: 0
Then I changed the TDM400P card to a different PCI slot and I get:
b2:/usr/src/zaptel# cat /proc/interrupts
CPU0
0: 6966 XT-PIC timer
1: 2 XT-PIC keyboard
2: 0 XT-PIC cascade
4: 321 XT-PIC eth0
9: 6143 XT-PIC wctdm
14: 1099 XT-PIC ide0
15: 2 XT-PIC ide1
NMI: 0
LOC: 6929
ERR: 0
MIS: 0
But then I found other evidence of sharing (which I verified with lspci -vv):
May 18 22:11:31 b2 kernel: Zapata Telephony Interface Registered on major 196
May 18 22:11:33 b2 kernel: PCI: Found IRQ 9 for device 01:07.0
May 18 22:11:33 b2 kernel: PCI: Sharing IRQ 9 with 00:02.0
May 18 22:11:33 b2 kernel: Freshmaker version: 71
May 18 22:11:33 b2 kernel: Freshmaker passed register test
May 18 22:11:33 b2 kernel: Module 0: Installed -- AUTO FXO (FCC mode)
May 18 22:11:33 b2 kernel: Module 1: Not installed
May 18 22:11:33 b2 kernel: Module 2: Installed -- AUTO FXS/DPO
May 18 22:11:33 b2 kernel: Module 3: Not installed
May 18 22:11:33 b2 kernel: Found a Wildcard TDM: Wildcard TDM400P REV E/F (4 modules)
May 18 22:11:33 b2 kernel: Registered tone zone 0 (United States / North America)
The BIOS always showed that I was sharing an IRQ with something else regardless of the
which of the four slots I tried.
I ended up with the BIOS telling me I was sharing with the USB; I turned USB off
in the BIOS along with everything else like serial and parallel ports.:
b2:/usr/src/zaptel# cat /proc/interrupts
CPU0
0: 24659 XT-PIC timer
1: 2 XT-PIC keyboard
2: 0 XT-PIC cascade
3: 205151 XT-PIC wctdm
4: 569 XT-PIC eth0
14: 1134 XT-PIC ide0
15: 2 XT-PIC ide1
NMI: 0
LOC: 24621
ERR: 0
MIS: 0
May 18 23:28:34 b2 kernel: Zapata Telephony Interface Registered on major 196
May 18 23:28:36 b2 kernel: PCI: Found IRQ 3 for device 01:0a.0
May 18 23:28:36 b2 kernel: Freshmaker version: 71
May 18 23:28:36 b2 kernel: Freshmaker passed register test
May 18 23:28:36 b2 kernel: Module 0: Installed -- AUTO FXO (FCC mode)
May 18 23:28:36 b2 kernel: Module 1: Not installed
May 18 23:28:36 b2 kernel: Module 2: Installed -- AUTO FXS/DPO
May 18 23:28:36 b2 kernel: Module 3: Not installed
May 18 23:28:36 b2 kernel: Found a Wildcard TDM: Wildcard TDM400P REV E/F (4 modules)
May 18 23:28:36 b2 kernel: Registered tone zone 0 (United States / North America)
...and lspci -vv shows no sharing either.
The test numbers are quite similar to those of the stronger machine
described near the end of this post. The test numbers were the same for all the PCI
slots and IRQ combinations I tried.
b2:/usr/src/zaptel# ./zttest
Opened pseudo zap interface, measuring accuracy...
99.987793% 99.987793% 99.987793% 99.987793% 99.987793% 99.987793% 99.987793%
99.987793% 99.987793% 99.987793% 99.987793% 99.987793% 99.987793% 99.987793% 99.987793%
99.987793% 99.987793% 99.987793% 99.987793% 100.000000% 99.987793% 99.987793% 99.987793%
99.987793% 99.987793% 99.987793% 99.987793% 99.987793% 99.987793% 99.987793% 99.987793%
99.987793% 99.987793% 99.987793% 99.987793% 99.987793% 99.987793% 99.987793% 99.987793%
99.987793% 99.987793% 99.987793% 99.987793% 99.987793% 99.987793% 99.987793% 99.987793%
99.987793% 99.987793% 99.987793% 99.987793% 99.987793% 99.987793% 99.987793% 99.987793%
99.987793% 99.987793% 99.987793% 99.987793% 99.987793% 99.987793% 100.000000% 99.987793%
99.987793% 99.987793% 99.987793% 99.987793% 99.987793% 99.987793% 99.987793% 99.987793%
--- Results after 71 passes ---
Best: 100.000000 -- Worst: 99.987793 -- Average: 99.988137
My Debian Sid (update/upgraded 2 weeks ago) system is thus:
b2:/usr/src/zaptel# uname -a
Linux b2 2.4.26 #2 SMP Tue May 3 16:44:31 EDT 2005 i686 GNU/Linux
Home grown kernel taken from kernel.org.
b2:/usr/src/zaptel# cat /proc/cpuinfo
processor : 0
vendor_id : GenuineIntel
cpu family : 6
model : 8
model name : Pentium III (Coppermine)
stepping : 10
cpu MHz : 996.783
cache size : 256 KB
fdiv_bug : no
hlt_bug : no
f00f_bug : no
coma_bug : no
fpu : yes
fpu_exception : yes
cpuid level : 2
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 mmx fxsr sse
bogomips : 1985.74
b2:/usr/src/zaptel# cat /proc/meminfo
total: used: free: shared: buffers: cached:
Mem: 260403200 23506944 236896256 0 1105920 15663104
Swap: 509923328 0 509923328
MemTotal: 254300 kB
MemFree: 231344 kB
MemShared: 0 kB
Buffers: 1080 kB
Cached: 15296 kB
SwapCached: 0 kB
Active: 9948 kB
Inactive: 8336 kB
HighTotal: 0 kB
HighFree: 0 kB
LowTotal: 254300 kB
LowFree: 231344 kB
SwapTotal: 497972 kB
SwapFree: 497972 kB
b2:/usr/src/zaptel# hdparm /dev/hda
/dev/hda:
multcount = 16 (on)
IO_support = 0 (default 16-bit)
unmaskirq = 0 (off)
using_dma = 1 (on)
keepsettings = 0 (off)
readonly = 0 (off)
readahead = 8 (on)
geometry = 4865/255/63, sectors = 78165360, start = 0
>
> D.
>
> FFF Managed Technology Ltd
> 60 Cook St
> P.O. 6368 Wellesley St
> Auckland
> t +64 9 356 2911
> f +64 9 358 9070
> m +64 21 415 297
> w www.fff.co.nz
>
>
>
> Waldo Rubinstein wrote:
>
> >Damian,
> >
> >Thanks for your input. Hyperthreading is in fact enabled and now that
> >you mention this I will disable it.
> >
> >The reason I ask is because under some load (may be 40 simultaneous
> >calls), voice quality degrades. We have audio problems where one
> >party hears the other but not viceversa and then it all works fine.
> >It's random audio quality problems in general. During these cases,
> >I'm constantly running vmstat 1 and CPU utilization is always 85%+ idle.
> >
> >I will also look into setting the TE410P in its own IRQ. Do you know
> >how I can do that? Is that a motherboard BIOS setting or is it
> >something that needs to be done to the TE410P itself?
> >
> >Thanks,
> >Waldo
> >
> >On May 16, 2005, at 12:59 AM, Damian Funnell wrote:
> >
> >>Hi Waldo, it really depends on who you ask - Digium say that
> >>anything less than 99.99% is going to result in problems, but ours
> >>regularly runs at around 99.98% and we don't have any problems.
> >>
> >>One of our boxes was running at around 99.96% and we had major
> >>issues with the voice quality packing up from time to time. We
> >>disabled hyper threading and put the TDM400P on its own IRQ and the
> >>results came back up over 99.98% (haven't had any problems since).
> >>
> >>Do you have issues with your * box? If so then I would start
> >>worrying about zttest output (and thinking about disabling hyper
> >>threading on those dual Xeons), otherwise have a smile and a beer
> >>and pity us poor fools who have had problems due to poor results.
> >>
> >>Cheers,
> >>Damian.
> >>
> >>FFF Managed Technology Ltd
> >>60 Cook St
> >>P.O. 6368 Wellesley St
> >>Auckland
> >>t +64 9 356 2911
> >>f +64 9 358 9070
> >>m +64 21 415 297
> >>w www.fff.co.nz
> >>
> >>
> >>
> >>Waldo Rubinstein wrote:
> >>
> >>
> >>>I was browsing the applications developed in zaptel and came
> >>>across zttest.
> >>>
> >>>After I run it, I get the following:
> >>>
> >>>Opened pseudo zap interface, measuring accuracy...
> >>>99.975586% 99.987793% 99.987793% 99.987793% 99.987793% 100.000000%
> >>>99.987793%
> >>>99.987793% 99.987793% 99.987793% 99.987793% 99.987793% 99.987793%
> >>>99.987793% 99.975586%
> >>>99.987793% 99.987793% 99.987793% 99.987793% 99.987793% 99.987793%
> >>>99.987793% 99.987793%
> >>>99.987793% 99.987793% 99.987793% 99.987793% 99.987793% 99.987793%
> >>>99.987793% 99.987793%
> >>>99.987793% 99.987793% 99.987793% 99.987793% 99.987793% 99.987793%
> >>>99.987793% 99.987793%
> >>>99.987793% 99.987793% 99.987793% 99.987793% 99.987793% 99.987793%
> >>>99.987793% 99.987793%
> >>>100.000000% 99.987793% 99.987793% 99.987793% 99.987793% 99.987793%
> >>>99.987793% 99.987793%
> >>>99.987793% 99.987793%
> >>>--- Results after 57 passes ---
> >>>Best: 100.000000 -- Worst: 99.975586 -- Average: 99.987793
> >>>
> >>>What does this mean? Should I have expected to get 100% across the
> >>>board?
> >>>
> >>>This is from a TE410P running on Debian 2.6.11-1-686-smp on a dual
> >>>Xeon 2.4GHz server.
> >>>
> >>>Thanks,
> >>>Waldo
> >>>_______________________________________________
> >>>Asterisk-Users mailing list
> >>>Asterisk-Users at lists.digium.com
> >>>http://lists.digium.com/mailman/listinfo/asterisk-users
> >>>To UNSUBSCRIBE or update options visit:
> >>> 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
> >>To UNSUBSCRIBE or update options visit:
> >> 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
> >To UNSUBSCRIBE or update options visit:
> > 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
> To UNSUBSCRIBE or update options visit:
> http://lists.digium.com/mailman/listinfo/asterisk-users
>
>
>
More information about the asterisk-users
mailing list