Stale Task issues related to DSP's

[bigbadbillyc]

Hi All,

I recently upgraded a MCD 4.1 system to MiVoice 7. The upgrade seemed to go reasonably well, however after completion, I had a major Alarm. When I went to look at this alarm I saw:

Category Name Alarm Level Total Unavailable% / Faults Minor Threshold Major Threshold Critical Threshold
Stale Tasks Major 3 33% >0 30% 50%

I checked to see that Reboot Resource recovery was allowed and the alarms were not cleared this morning (should have occurred last night). I also tried performing a cold boot this morning and these did not clear (but did set up a fun new Resource Recovery Reboot for again tonight

PROGRAMMED REBOOT SCHEDULE IS OFF
PROGRAMMED REBOOT SCHEDULED: DAILY 02:15:00
PROGRAMMED REBOOT RESOURCE RECOVERY IS ON
PROGRAMMED REBOOT RESOURCE: RECOVERY 02:15:00 (PENDING)
PROGRAMMED REBOOT COURTESY IS OFF



Looking in the Maintenance Error Logs I can see the following:
* 2657 2014/Jun/26 15:40:42 Task Monitor Major Alarm being raised upon detecting the following task suspended: tDspBoot0003 0x453ba80
at: THU JUN 26 15:40:17 2014 Resource recovery reboot will occur if programmed
* 2655 2014/Jun/26 15:39:52 Task Monitor Major Alarm being raised upon detecting the following task suspended: tDspBoot0002 0x453bdc0
at: THU JUN 26 15:39:26 2014 Resource recovery reboot will occur if programmed
* 2650 2014/Jun/26 15:39:01 Task Monitor Major Alarm being raised upon detecting the following task suspended: tDspBoot0001 0x453d7c0
at: THU JUN 26 15:38:36 2014 Resource recovery reboot will occur if programmed
* 2647 2014/Jun/26 15:38:16 Task Monitor Major Alarm being raised upon detecting the following task suspended: tDspBoot0000 0x453db00
at: THU JUN 26 15:37:50 2014 Resource recovery reboot will occur if programmed

When I try to look at the DSP's themselves, I *think* they seem fine:
SHOWDSPSTATUS
Dsp Status: (Warning! This only displays the current state of the dsps in the system. Failures previous to a system reboot are not recorded.)

Slot Number 1 Dsp 0 State: IDLE
Slot Number 1 Dsp 1 State: IDLE
Slot Number 1 Dsp 2 State: IDLE
Slot Number 1 Dsp 3 State: IDLE

Slot Number 2 Dsp 0 State: IDLE
Slot Number 2 Dsp 1 State: IDLE
Slot Number 2 Dsp 2 State: IDLE
Slot Number 2 Dsp 3 State: IDLE

Slot Number 6 Dsp 0 State: ACTIVE
Slot Number 6 Dsp 1 State: ACTIVE
Slot Number 6 Dsp 2 State: ACTIVE
Slot Number 6 Dsp 3 State: ACTIVE


A DUMPDSPBOOTINFO shows the following:
Dsp(0,0): mmcSlot=2 (1-based) Load Type: TRANSCODER(2,"/sysro/DspConfig/TRANSCODER.161")
Mod Type: DEFAULT(0)
Dsp(0,1): mmcSlot=2 (1-based) Load Type: TRANSCODER(2,"/sysro/DspConfig/TRANSCODER.161")
Mod Type: DEFAULT(0)
Dsp(0,2): mmcSlot=2 (1-based) Load Type: TRANSCODER(2,"/sysro/DspConfig/TRANSCODER.161")
Mod Type: DEFAULT(0)
Dsp(0,3): mmcSlot=2 (1-based) Load Type: TRANSCODER(2,"/sysro/DspConfig/TRANSCODER.161")
Mod Type: DEFAULT(0)
Dsp(1,0): mmcSlot=3 (1-based) Load Type: TRANSCODER(2,"/sysro/DspConfig/TRANSCODER.161")
Mod Type: DEFAULT(0)
Dsp(1,1): mmcSlot=3 (1-based) Load Type: TRANSCODER(2,"/sysro/DspConfig/TRANSCODER.161")
Mod Type: DEFAULT(0)
Dsp(1,2): mmcSlot=3 (1-based) Load Type: TRANSCODER(2,"/sysro/DspConfig/TRANSCODER.161")
Mod Type: DEFAULT(0)
Dsp(1,3): mmcSlot=3 (1-based) Load Type: TRANSCODER(2,"/sysro/DspConfig/TRANSCODER.161")
Mod Type: DEFAULT(0)
Dsp(2,0): mmcSlot=7 (1-based) Load Type: TELEEC(6,"/sysro/DspConfig/TeleEC.363")
Mod Type: NOCONFERENCE(1)
Dsp(2,1): mmcSlot=7 (1-based) Load Type: TELEEC(6,"/sysro/DspConfig/TeleEC.363")
Mod Type: NOCONFERENCE(1)
Dsp(2,2): mmcSlot=7 (1-based) Load Type: TELEEC(6,"/sysro/DspConfig/TeleEC.363")
Mod Type: NOCONFERENCE(1)
Dsp(2,3): mmcSlot=7 (1-based) Load Type: TELEEC(6,"/sysro/DspConfig/TeleEC.363")
Mod Type: NOVOICEMAIL(7)

Any ideas why I can't get rid of these Stale Tasks Alarms or what problems they might cause? I hate to see a system with a Major Alarm, but at the moment, the system seems functional.

 

[Lundah]

Could be a flaky DSP module. You may want to try removing the 161 MMC's one at a time and then bring the controller back up to see if the remaining DSP's load with no issues. Unfortunately the 363 type DSP in slot 7 is embedded on the controller midplane itself, and isn't field replaceable. If that doesn't work I'd be opening a ticket with Mitel Tech support.

 

[bigbadbillyc]

One of the two SHARC 21161 DSP's seems to be the culprit. After removing it, the system seemed to recover and get by without these errors. I had stuck it in the system a long time ago as a transplant from one of our older 3300 controllers and it is a hardware REV A board (not sure if that is a contributing factor to the failure or not.) In any case, I didn't truly need all 64 compression channels I had programmed, so I bumped down the number I was attempting to use in the System Capacity Selection form, and am now running without error with just the one 21161 installed.