[asterisk-ss7] chan_ss7 MTP is now Down after 70 calls.
Robert Thomas
thomcr at gmail.com
Tue Jun 28 03:31:01 CDT 2011
CPU: Dual Intel(R) Xeon(R) CPU X5650 @ 2.67GHz
2 Times ( CPU has 6 cores)
Memory: 3 GB
ASUS Z8PE-D12X – Motherboard
TE410P
Load is never above the 0.5 out of 24 available.
Asterisk takes like 130% of CPU time.
About interrupts I'll check that during load hours. This is a test run
running right now.
root at ostional:~# dahdi_test
Opened pseudo dahdi interface, measuring accuracy...
99.986% 99.985% 99.984% 99.980% 99.982% 99.987% 99.983% 99.981%
99.987% 99.983% 99.983% 99.984% 99.982% 99.984% 99.984% 99.983%
99.981% 99.984% 99.985% 99.986% 99.978% 99.984% 99.986% 99.979% ^C
--- Results after 24 passes ---
Best: 99.987 -- Worst: 99.978 -- Average: 99.983408, Difference: 100.016591
A look at the proc interrupts now
root at ostional:~# cat /proc/interrupts
CPU0 CPU1 CPU2 CPU3 CPU4
CPU5 CPU6 CPU7 CPU8 CPU9 CPU10 CPU11
CPU12 CPU13 CPU14 CPU15 CPU16 CPU17 CPU18
CPU19 CPU20 CPU21 CPU22 CPU23
0: 33 0 0 0 0
0 0 0 0 0 0 0
0 0 0 0 0 0 0
0 0 0 0 0 IO-APIC-edge timer
1: 2 0 0 0 0
0 0 0 0 0 0 0
0 0 0 0 0 0 0
0 0 0 0 0 IO-APIC-edge i8042
8: 1 0 0 0 0
0 0 0 0 0 0 0
0 0 0 0 0 0 0
0 0 0 0 0 IO-APIC-edge rtc0
9: 0 0 0 0 0
0 0 0 0 0 0 0
0 0 0 0 0 0 0
0 0 0 0 0 IO-APIC-fasteoi acpi
12: 4 0 0 0 0
0 0 0 0 0 0 0
0 0 0 0 0 0 0
0 0 0 0 0 IO-APIC-edge i8042
16: 11246 0 0 0 0
0 0 0 9998 0 0 0
0 0 0 0 0 0 0
0 1856235 0 0 0 IO-APIC-fasteoi
uhci_hcd:usb3, uhci_hcd:usb8, pata_it8213, wct4xxp
17: 11244 0 0 9998 0
0 0 0 0 0 0 0
0 0 0 1856225 0 0 0
0 0 0 0 0 IO-APIC-fasteoi wct4xxp
18: 66 0 0 0 0
0 0 0 0 0 0 0
0
On Tue, Jun 28, 2011 at 3:18 AM, Krzysztof Drewicz <
krzysztofdrewicz at gmail.com> wrote:
>
>
> 2011/6/28 Robert Thomas <thomcr at gmail.com>
>
>> Hello Everyone,
>>
>>
>> We have a 7 E1 SS7 linkset that above 70 calls load starts to flap
>> constantly. The MTP goes down, and comes back after 10 seconds or so.
>>
>
>
>> Could this because of line errors, slips or problems on the line? How
>> could we check that from the asterisk side?
>>
>
>
> Too much load on asterisk box? If you have Digum's card there is a tool
> that checks how accurately hardware interrupts are being handled / routed
> 1st check the /proc/interrupts file,
> 2nd look for tool named zttest it should yeald at least 99.98 %
>
> This is no slip problem, as if i got it correct: if you don't run any
> calls, your digital lines are stable (no alerts etc...)
>
> What kind of hw you have there ?
>
> --
> _____________________________________________________________________
> -- Bandwidth and Colocation Provided by http://www.api-digital.com --
>
> asterisk-ss7 mailing list
> To UNSUBSCRIBE or update options visit:
> http://lists.digium.com/mailman/listinfo/asterisk-ss7
>
--
Robert
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.digium.com/pipermail/asterisk-ss7/attachments/20110628/7550b334/attachment.htm>
More information about the asterisk-ss7
mailing list