Power Cycle on 81C (Software 3.0)

Judas Priest!!

No offense Ralph, but horror stories like yours serve to periodically rem>I'm planning to power cycle my 81C to perform a reload on the controllers

If I have not mentioned it lately, bite my big beef lollipop :-)

Can I truthfully say that the Nortel runs better than the Mitel??...not a chance because the 64 Mitel systems left in the US are run by the best of the 8,400 techs that supported them in their glory days.

The Meridians are so plentiful, a significant number of them are managed by people who have less experience then required to make them shine.

The Mitel is like a Checker car (started once per day, turned off once per day...they filled the 350 Chevy engine crankcase with gear oil) while the meridian is like the Toyota that can run for years with the factory oil.

If you took the better of Mitel admins and the better of Nortel admins and sat them in a switch room with their system, they would both be stuck with equally boring jobs for many years.

Gene from, among other domains -

Seeya

PS - IPC has a remote monitoring service now (call for help system) Is that really new or did our reps just decide to tell me it was new?

OK, a serious answer. As far as powering the system off. If it is a DC system, I wouldn't power off the shelves. Just turn off the 5 breakers on the back of the pedestal. A few years ago, Nortel make some bad power supplies if you turned off the switch on the front of the power supply it would break internally. After that happened a couple of times, I always used the pedestal breakers. It's bad power supplies are not a problem if you have a recent switch. On a AC switch, it really doesn't matter how you turn it off. Powering back up, I would power back up 1 column at a time, one after the other, to make sure I didn't cause a surge. just make sure they're all backup before the sysload completes. If you're just wanting to download all controllers, You can do that by going to ld 97. Do a chg type sysp cr to fdlc and type in all f l (all forced latest) and cr thru. This change is only good until you init the switch and than it will revert back to what it was. When you init the switch it forces downloads to all cards that require software downloads such as controllers and msdl cards. This init can be rather long because of this. It is usually done only after a software upgrade to force the software out to such cards, but can be used anytime. However be aware if there are marginal card out there, you need to have replacements ready.

As far as cold turn offs to correct data base problems. I prefer to do a parallel reload, if there is a problem on sysload, it is easy to back out and correct the problem. If you are going to do a cold restart, make sure you print all the stations first. Watch your sysload messages to see if it blows away any of your tn's. It's goes without saying to make sure you have backups available if something goes wrong. I would do one on core 0 and than switch processors and do one on core 1. I have seen a backup on one core be defective whereas the other one worked perfectly.

"Ralph Shapiro" wrote in message news:zG7be.8$ snipped-for-privacy@okeanos.csu.net...

Ours are all DC-powered and since 1986 I can recall there having been only two "power-down" events; once to move a system from Kansas City to Houston and once to perform a forklift upgrade (from 5 cabinets of the old SG architecture to nearly 40-some cabinets of Fiber-Distributed SX2K Light). Can anyone say "cluster"? Yes, I'm very much aware that "cluster" is only half a word. (Clint Eastwood/Heartbreak Ridge)

Still power-cycling seems drastic, like killing a fly with a sledge-hammer. Gee whiz, isn't there some other way?

controllers

If your 81C is like mine - just wait a while and it'll probably power cycle itself.

We can pretty much depend on our switch going down at least once per year - redundant, UPS, and a diesel generator behind that - it just likes to die.

Of course we lose all logs so there's no way to tell what caused it.

NEVER had this problem with Avaya switches!!!!

I realize this is a delayed follow-up, but what do you think the problem is? If it is only once a year or so, I'd guess you mean it just flakes out - not really a power issue, but more like it can't handle something and just gives up. We'd be interested in the software release. But if I were to guess, I'd say they screwed up grounding and when they hose down and wax the substation, you pay the price? (kidding with the specifics of course)

Do you have a UPS for the PBX, then a larger one before the Generator? Two in series, so to speak?...if so, that can cause havoc, usually the dumber of the two UPSs pays the price and - decides to act up. Won't happen daily, but will at times

If you have a mongoloid Option, this is a place to tell it to the world. I assure you Nortel is reading... but better yet, someone might respond with ammunition.

Actually, you do NOT lose all the logs. There is a log that survives sysload - two, actually. Unless you were to completely dump power to the switch, both sides shouldn't just die. Some questions:

What release? What processor? (all that will be in the issp mentioned below) Is it AC or DC powered? Probably AC, from the comment about UPS. If you have voicemail, does the voicemail restart at the same time? What do you need to do to recover? Does it just come back up?

Have you been there when it happens? What's in the history when it comes back up? Just a sysload?

How well does it run when it's not screwing up? Do you have bug, and err turned on? What's in it for patches. Ld 22, and do an ISSP and email it to me. I need everything from the Req: issp to Req: The bobsjunkmail address is real - just put something obvious in the subject line. What does your vendor say about it?

If it's reloading, and the voicemail is booting at the same time, your UPS is probably hiccupping. People talk about grounding issues causing problems. Even a complete absence of a ground won't make both sides reboot sporadically at the same time, while the rest of the time it runs normally.

One thing is for sure: what you are describing is abnormal, and should not be tolerated. I have high call volume switches that run YEARS without so much as an initialization.

Release 3 - with supposidly all the patches.

We only recently added the diesel generator. System crashes every once in a while - 4 times in the last 3 years. It just hangs - no dialtone, no nothing. It takes a reboot to get it back on line.

Used to be Nextira and now Verizon just shake their head and say, "Well, its an electronic device and you have to accept that it will crash."

I just don't buy that. Since I rely on Channel Partners for technical advice I'm a little bit lost.

Its VERY frustrating to say the least.