Version 9.0 RTR

[McGruff753]

Anyone out there having problems with their ACD RTR (Rolling To Redundant) that recently upgraded to Version 9.0? My ACD has lost its Primary 3 times in 3 weeks. We have already replaced the Utility Card in the Primary. Now we may be looking at the Processor or the Hard Drive. Just curious if anyone else out there is seeing similar issues. Thanks.

 

[chris13]

We upgraded 2 months ago, no RTR / CTB until now.

RTR is always the symptom, not the problem itself. The software switches to backup to prevent further impact to your business...
Any hints from the logs or aspects support about the situation before the RTR?

 

[McGruff753]

Yes, I realize it's not the root cause. I was looking to see if anyone had a similar issue and if they knew which hardware item may have been the cause, because Aspect can't find anything in the logs. As I stated, we replaced the utility card after the first RTR and it still RTR'ed again. At Aspect's request, we just replaced the primary processor, but they say it could still be a hard drive.

 

[chris13]

I asked 2 other people who are running Rel.9 now for some weeks/months, but none of them have experienced similar problems, too.

Last year we had issues with RTR/CTB where Aspect couldn't find the cause at first. But that was running 8.3.
They replaced a shelf-controller and even the mainboard of the primary server without success.
The problem disappeared after all(!) shelf-controller cards where replaced by new ones which were of the same revision level/version...

 

[McGruff753]

Cool, I'll keep that in mind. I know the SSC's and SMC's are now being phased out, but we were told we didn't need the new SMC2's until 9.1. If it happens again, I'll suggest we make the change to the new SMC2's and see if that helps the issue.

 

[chris13]

Murphy must be watching us...
Friday wired messages appeared in the log of the backup-system. For my understanding they were generated by the routines wich regularely check for leaked call-registers, correct assignment of adresses and their usage when the "3000 Ports" option is enabled. The taxi-ring lost its base-address and other config... During the night both controllers were shutdown and shelfs powered off. Today everything seemed to be fine, but just when I wanted to leave my desk the backup-controller decided to reboot without any visible cause.
It's months ago that we upgraded to Rel.9. New CPU, a 4th shelf was added and they installed an IP-Card for VoIP/Uniphi Connect...
Too bad, a year ago I was transfered to another department which just covers the configuration/call-routing. My ex-collegues -which are responsible for the hardware, power, cooling and trunks- have no real clue what's going within the black-box. Really BAD situation... when I was responsible for the whole thing, I worked together with chief of the german support and digged deep in the bits and bytes of voice and networking/software -I know both very well. But now I have to call somebody to call Aspect...
Sorry, probably that should be a start of a new thread in

http://www.tek-tips.com/threadminder.cfm?pid=655,

but I couldn't resist.
Perhaps our backup-controller has the same problem as your primary, we'll see...

 

[McGruff753]

Let me know if they figure it out. Good luck.

 

[chris13]

Thanks Aspect event didn't suggest a reboot 'cause they want to do some further analysis...
As long as I can't help them, I don't mind not beeing the one to be on call this week...
I'll keep you up to date.

 

[chris13]

Oups... my brain was near suspend-mode and didn't recognize the difference between RTR and CTB. The problems last year were CTB (switching to backup), not RTR (backup dies and comes up again).
The situation yesterday was a lucky combination: Backup died, then primary wanted to switch to backup but it was already standalone. That way processing and everything else was not interrupted except no voice system for some minutes.
The problems seem to be caused by the new shelf or the IP-card within. Counting all together that makes good sense, I personally suspect somthing concerning its backup controller or the cabling of the backup-stuff within the shelf. We'll see...

 

[McGruff753]

Hmm, according to my previous Aspect dealings, and how Aspect opened these last 3 tickets, I think they look at CTB (Cut To Back-Up) and RTR (Roll To Redundant) as the same thing nowadays. I think when the back-up dies, or reboots, they call it something else. They seem to always change their acronyms just when I get used to the previous ones. We're still stable here (I may have just jinxed it) so far. If it happens again, we may need the hard drive or change out the SSC's. We've got 8 of them at $2000 a piece to relace. Not in the budget this year. Come one baby, hold together.

 

[chris13]

Good luck...
"...hold together" - until when? New budget available?
Today I was told that during the upgrade to Rel.9 we got 8 of the newest shelf controllers to compensate multiple downtimes caused by the problems last year.
Probably it didn't hurt them too much 'cause the upgrade was bundled with other orders (4th shelf, Uniphi stuff like IP trunk card, proxy, clients etc) and everybody knew th rumours about an upcoming project concerning CTI/IVR/CRM...

Something completely different: Aspect never was the cheapest offer when it comes to support contracts. But I think that there is a difference in what you get for the money since 1997. In the old days I was really impressed by their commitment to customer service. One day I called the hotline just to ask something about CustomView, and got an anouncement telling me that all their people were busy with other high priority cases and that I could leave a message, expect some a longer time in queue than usual or decide to signal an emergency, too. I dropped the call. Half an hour later I got a call from Aspect. The technician told me that he saw that I was in their queue and asked if he could me in any way. I was really surprised and delighted
Times have changed, business has changed and everybody has to save its money wherever it is possible.
But I don't like to be promised that heaven would come to earth if we just ordered the next release. Not when I can't get any info on what caused our problems in detail and how it is fixed.

Tonight some paddle boards and probably the new shelf are going to be replaced.
But our box couldn't hold together until evening. When the bells rang noontime no more incoming calls where accepted, but no problem with outgoing calls. Same problem with the ISDN-stack (Trillium) like a year ago when we were on 8.3.

Murphy seems to have a very good time now...

 

[McGruff753]

Well, if it could hold together until July 1st, I would be happy. My last day here is June 30th before I move to another Aspect ACD customer site. But the manager here is going to put it in the budget for next year. They had budgetted for an IVS but it the money got used for other things. It's a good thing, too, since 9.1 eliminates the need for an IVS.

I know what you mean about the service contracts. I think we pay over $120,000 a year just for this one site. (We have 3 total)

Do you have Custom View? And if so, have you noticed the flaw at the top of the hour in the voice system? Aspect thought they fixed it with Version 9 but it is still there. Ever since we went to version 8 four years ago, our voice system shows 85% full on the hour until 5 minutes past the hour. Then the reading is more acurrate. (Like 15 to 20%) Now that we're on version 9, it shows 96% full on the hour until 5 minutes past the hour and corrects itself to 10 or 11%.