CICS dropping calls....I think

[moonltgmblr]

System description CICS rel. 1.0 sw with Startalk Flash 1.7. Ok, on first service call I am told that system has been in place and working fine for a long time...this is a new problem. When calls are put on hold they are getting disconnected. It sounded to me like a definite system problem. I swapped out the 2 LS/DS cards. Problem continued. Oh, by the way, this is the problem presented, not duplicated by me...but how could the user screw up putting a call on hold? Then on one occasion I called in and call was dropped after someone had attempted to put the call on hold. I heard what I might describe as a very quick splash of static. OK, now I believe there is in fact a real problem. I have an MICS with old software so I do a swap. I am told the problem is still there. AT&T assures me, they have changed out copper and equipment. I am also being told by the customer that periodically 911 is being called and showing up at their door. First thought is something is tapped into their line but from equipment to mpoe it is not tapped off to any devices. I have tried much more and can go into it futher if anybody wants to get involved and think they can help...I would welcome your imput.

 

[Crowtalks]

Since you swapped the system out and the problem is still there, I would start swapping out phones, keeping note of which ones you have changed out, until you find which one is causing the issue.

As far as phantom 911 calls, if it happens again, ask the customer to check, or you can check, and see which trunk dialed 911. 95% of the phantom 911 calls I have encountered in twenty five years since we have had 911 in our area relate to analog static in a cable pair, jack or block. look for corrosion in or around the demark or in the amp cables.

NCSS NCTS NCTE

 

[moonltgmblr]

I am still at it...have swapped out phones (one at a time) and I am being told that the problem still exists...calls dropping on hold...and on at least two occasions, while call was connected. '911' calls continuing, I changed cable pair from main distribution frame to intermediate distribution frame but not from that idf to idf where equipment is located. House cable, I really doubt cable problem. Also, I once again, changed out phone system in case I might have missed something the first time. I am wondering about design of these circuits...something I should know about...as far as the Norstar is concerned not a whole lot to setting up loop start trunks. Thank you for chiming in, a am sorry it has taken me so long to get back to you. I am appreciative.

 

[Crowtalks]

Probably won't matter a hill of beans...make sure lines are supervised (yes)

NCSS NCTS NCTE

 

[oldestgeek]

Agree with Crowtalks - the problem is likely static on the line. Have the customer request the Telco come out and do a loop current, silent termination, and tone test. I would even go and on a limb and bet that the trouble happens more frequently when it rains.

 

[Crowtalks]

oldestgeek, huh...I'm 54 and a geek

NCSS NCTS NCTE

 

[Mountainbear]

Sounds like analog problems, therefor it would most likely be before the system. I can't see a phone set causing this issue, but check any system / event logs and clear them so any new ones will be the only ones in the system. Other than that find out what number is calling 911, listen to that line then pull everything off that line, (one at a time) from the inside (after the demarc). Listen for static, crackle, or any thing other than dial tone, (I dial half a number to get silence). While you have the line pulled you might as well look at the inside wire with a multi meter, you should see infinit resistance or very close from tip to ring and both sides to ground (not sure through a phone system). Also check for green at the demarc. Good luck. As mentioned you may have more luck on a wet day.

 

[moonltgmblr]

Well, I wish I could tell you that the problem has been resolved...but I can't.
Let me see...since we last spoke I have changed out the ksu twice, changed out phones six at a time. And lastly, changed out the voicemail...still the same problem. The local phone company told me early on that the copper and equipment had been changed out. That to me was all they could do. Today, I am given, maybe unintentionally, but given the impression that that may not have been done. And that was being done today. I will just have to wait and see...this is one of those jobs that is classified as an effort in futility.
Some people are Ford people, others Chevy people...some Avaya....me, I am Nortel. Can't believe Nortel at fault here.

 

[Mountainbear]

Changed the KSU, twice AND the voice mail? You have certainly gone beyond the call of duty.
Having worked for two different service providers (1 cable, 1 copper) I can be fairly certain that they did not replace the copper or equipment. To a phone company, if they cannot physically see a problem with their test equipment then the problem is CPE (customer provided Equipment), even if they won't say that to you. What they may have done is moved them to a different pair from the C.O. In fact the big telco's almost assume it it CPE unless they see a problem or a trend with multiple customers.

If it is one line causing problems can you move this on the switch? Or even make it analog to a few locations, not tied into their switch at all? I know that seems drastic.

The other option even more drastic is to move that number to a different provider, Cable if possible. If nothing else that will get AT&T's attention.

 

[gforce9]

Mountainbear is correct; Telco only does Class A inspections (changing cable pairs, hardware,.....) when crap has hit the fan (somebody went 'presidential'). It is a cable problem outside, most likely, and they need to find it. Static/ground hum/noise may or may not be present with cable issues. An SMDR device goes a long way in cases like this; both for isolating a bad line and false 911's. Have them write down the time, event, station #,....when a call gets cut off, then go back into the SMDR records and start building your case. You have done everything possible......and then some. In a VERY nice way, restate that to the customer; they tend to run thin on patience, forget what has already been done, and don't really care who's at fault, they just want it fixed. Get an SMDR and connect it to one of their pc's that can stay on 24/7, open hyperterm or ProComm and let it run!

 

[Crowtalks]

Yes you can move it in the switch. Let's assume it's the 1st number in the telco hunt and you have 4 lines coming into the Compact...swap lines One and Four.

At this point when people call that published number (the first line in hunt) it will ring to line four position in the switch because that's where you moved it to. You can swap line appearances on the phones on a per station basis in the switch by way of a feature code. I think it's F 81 or F *81 (If I'm incorrect on the feature, someone will catch it and correct me. You go to each affected phone and swap line appearance 1 and 4 on the phones and. as far as the customer is concerned, line 1 is still assigned to line 1 (even though it's actually line 4 now).

If you discover the problem moves with the line, get the telco to swap cable pairs, or do further testing.

NCSS NCTS NCTE

 

[jerryreeve]

you swapped the KSU out twice. how about the line cards?

the suggestion from gforce9 is very valid and instead of the pen and paper you can use account codes from the phone that tried to put the call on hold and that will have the call itself flagged on your print record or file. I have used this method and works OK with some customers and others prefer the pen and paper method.

----------------------------
Hill?? What hill??
I didn't see any $%@#(*$ Hill!!
----------------------------
JerryReeve
Communication Systems Int'l
com-sys.com

 

[moonltgmblr]

The way this problem was presented to me is that it is multiple lines that are dropping calls. The local phone company just changed out equipment for first three of six lines. The problem is still happening. I in an attempt to eliminate all variables on customer side, changed cable pairs on all lines from minimum point of entry to equipment (including the two 25 pair cables that connect system to backboard). In the process I found one line (line 2) that at first intermediate frame took a dogleg off to a 21x demarc device and then on to house cable to second intermediate frame. I bypassed the RJ21X to elimate a few of the many splice points and am now in a wait and see. Only thing left on my side are the individual station cables. Thanks again for all the help.

 

[gforce9]

These issues are always tough to find....sometimes REALLY tough. I doubt it is individual phones, unless it is always the same phone. T-Taps could be it and you do need to eliminate those if there are any. Keep this in mind; there is, potentially, 20x more equipment in the Network than there is on a particular customers site. It could be that these people getting cut off are all riding a DS3 out of a city 300 miles from the customers site. No testing (or equipment change) will affect or demonstrate that 300 miles away. It could be a local DS1/3 issue out of that local office before the dialtone hits the copper in the SAC. Did Network look into that?. With that, what eq did the C.O. really change? A jumper? An analog card in the switch? There are lots of variables and only diligence and time will reveal the issue. Here are some thoughts:
1- Honestly, I believe you have a Network issue.....somewhere.
2- You have already changed all the hardware, correct? Trunk cards, too? If there is nothing left to change, stop....the problem is elsewhere.
3- Double check for t-taps (as you have) from the demarc to the KSU and consider running a feeder if there is any doubt to its integrity or runs right next to a new power source.....in other words, be a good detective and ask lots of questions of the maintenance guy.

Here is what I would do:
1- Find out where these calls are coming from. Get an SMDR on that system.
2- Have them document time, date, incoming line/trunk # the problem happened on, hopefully, the calling # (what # initiated the call), and a description of the problem. See also Jerry's idea a few posts up.
3- Find out what Telco actually did (this will be like pulling teeth, but you need to know EXACTLY what they did)! They may be lying to you. They may have put the order through, but the work hasn't really been done yet. Someone may have changed a single jumper in the C.O. and called it a day. You MUST find out.
4- Make sure you keep everyone (including and especially you're boss) updated on what you have done. The last thing you want is an angry, uninformed customer blindsiding your boss. This comes from personal experience.
Happy hunting!

 

[curlycord]

What is the NT # on the line cards are you using.



=----(((((((((()----=
curlycord

http://www.curlycord.com

 

[moonltgmblr]

I have thought of setting up smdr to track bogus 911 calls but do not have the equipment available for testing.
Got a call today from office administrator and she told me there was one reported dropped call on line 1 but other than that it seemed pretty much well behaved. When they (AT&T) tells me the equipment has been changed. I have to believe them. I was not happy when I found multiple splice points on line 2 taking the line through RJ21X...that's a 50 pair connector, through bridging clips (old) and cross connected back to house cable. Lot's of places for that static or poor connection, possibly reducing line current....I don't know, as I told the customer...I am grasping at straws.
Oh by the way, the top dog at this company suggested Feng Shui.

 

[jerryreeve]

Ah yes, align the CICS in a north/south direction with the front facing the door to the equipment cubby. and make sure that the cross connect blocks are east/west facing the d-mark, burn cross connect insulation as your incense stick in the middle of the cubby for the last 2 hours of daylight on the next equinox.

----------------------------
Hill?? What hill??
I didn't see any $%@#(*$ Hill!!
----------------------------
JerryReeve
Communication Systems Int'l
com-sys.com

 

[moonltgmblr]

 

[Crowtalks]

'Burn cross-connect insulation'...are you sure?
I thought you only burnt cross-connect insulation at the anointing rite of the virgin line card...

NCSS NCTS NCTE