sillejo

Bill,
There are a few people on here that have done this or are going through this, but as the moderator of the Cisco section, I will tell you that a better place for this discussion is the Nortel forum and would encourage you to copy the post there. The guys answering questions still have mostly Nortel experience and therefore look at those forums more often than they look here.

Secondly, I see a lot of people come to the forum with pre-conceived ideas of how Cisco products are implemented or should be implemented and many times these assumptions come from interaction with just the Cisco reps or from inexperienced IP Tel engineers (read: CCIE's that are green to how telecom really works) If you are connecting with MPLS then the remote "flavors" of Cisco should realy be considered as Disaster Recovery solutions and less as remote solutions. Right now I am helping a global client with facilities in EU, SA, AP and NA. They do this with 3 clusters (EU, SA, NA) and 3 Unity Vm solutions. If there is a good partnership with the global carrier (including link monitoring and SLA's that include VoIP tags with QoS) then there should be no need to distribute the Administration of the voice management to the routers (SRST or CME, either "flavor") What should be planned is the DR of those sites. CME is going to give the site many more features than SRST and it can give you a GUI interface to manage it (convenient if you have some kind of backup data link not suitable for voice) Local voicemail will be the biggest thing to overcome in a DR situation and I would suggest that CUE or Unity Connection be purchased for on-site backup and configured with general mailboxes for use during the DR time.

No matter which way you go, the remotes should primarily run off of the CM clusters and not off of the remote solutions. Stories of this being a bad solution are rampant and that is due to an overabundant faith in QoS over Frame networks back in the early 2000's These were over pitched by Cisco and providers a like and were destined to provide poor quality service. The move to MPLS has all but eliminated 95+% of all the issues experienced with the old Frame set-ups. I would say the top issues are Expense for local conference resources, VM during DR and faith in MPLS provider (this last one is key. weather it is Verizon, AT&T, BT, or Orange, get them to give you a reference that uses MPLS in the same part of the world as you will. Location Location Location - for instance, Part of Africa and Latin America offer "MPLS" but it doesn't matter if your round trip jitter is 400ms)

good luck, post your questions and remember this: "features" are a commodity and all the old telecom vendors like Nortel, Avaya, Seimens, NEC all "swear" Cisco does not do X and almost every time it's because they have 2 year old software in their lab or simply don't know what it's called outside of their world. Changing phone systems is the perfect time to get rid of all the crap overhead that has been implemented then forgotten and left unused over the years (mostly I am thinking of pickup groups and shared lines that would be better served with small ACD's and a Presence application) Users are smarter than they seem (accepting change) and dumb as posts (remembering the changes after vacations), but don't get sucked into a like for like RFP bidding war. Everyone will tell you that they can do it and then you will be stuck with the crappy results or overbudget implementations. Let the vendors re-do flows and set-ups where appropriate and budget for station/department interviews so whoever puts in their system doesn't wast a bunch of time re-creating features that the user no longer needs or uses. You will be shocked when you see what is learned, what you knew, what you didn't know and what you thought you knew.

__________________
Deus ex machina

aikenmarvin

I'll follow above with the statement of:
We went with a supposedly VAR that Cisco rep suggested. These guys are mostly Cisco data guys. They don't know telephony all that well. The BIG problem I have so far is that the VAR doesn't know the product as well as they should. We are already over budget, and calling them on the carpet for half of the overruns. THEY have been given our configs, coming back to Us and saying, " Oh if you want that then it will cost extra..." The Cisco sets will not do everything that Nortel will, the Unity voice mail needs 2 pieces, the Unity and IPCC to do everything that the Call Pilot will do.

IF you want a good intergration, then go with a BIG VAR, such as SBC. AND remeber if they quote a complete Cisco system that is less than a comparable Nortel,,, their DEAD wrong. The Cisco will cost you about 25% more than a Nortel or Avaya.

sillejo

Some of that is fair criticism, but some of it is unfounded:
For instance many of the lager implementors like SBC have one or two guys then outsource the work to independent contractors. They have recruiters calling all the time for 6-9 month gigs (Verizon, Berbee, SBC, etc.) If the Cisco rep recommended them, then it means you didn't have an existing relationship with a Cisco vendor, which also means you are part of the 5% of the world running something besides a Cisco backbone. You got the partner of the month (whoever they think is up and coming and needs the business) A better route would have been to bring in 2-3 competing Cisco vendors and take the Cisco rep out of the mix. Once the projects starts the Cisco guys will be hard to find and you will be lucky if you seen them again before it's over. The Partner is key as their guys wil l be the ones conducting the interviews and configuring the software. I would also guess that the Cisco guys put together the Bill o' Materials and they did this before anyone did any investigations. This is a classic problem: Cisco comes in and makes the client think they are getting a bunch of free advice and consulting by doing drawings and a preliminary BOM, then they introduce a partner who is not experienced in pushing back on Cisco. THey accept the BOM "as is" then find out later that the client had talked to Cisco about a bunch of advanced feaures that are "possible" but not with the BOM they are getting, thus the added features and upgrades.

If it's scoped correctly from the beginning and the Partner is involved from the beginning (not given a BOM and told to put it in), then the cost overrruns never materialize and the feeling of being upgraded to death does not happen either. It also is true that the Cisco system will cost more, but it's not 25%. It is typically the premiun cost associated with hardware (when a company controls 90+ percent of a market) which will come out around 8-12% more. This additional cost up front is what Cisco argues will be re-couped within the first two years by having a single vendor across the enterprise. This again goes back to execution. If the client does not follow through and starts integrating with an old Octel here and a DMS there, and a Mitel at the other place, then that 8-12 never gets recouped.

Now, as for the core decision. It sounds like the last post was a situation where price was the main factor from the telecom side of the house. If price is your main concern then the short term solution is more important and you should never, ever look at a Cisco or Avaya pure IP system (these two own over most of the pure IP systems out there today...sorry, Nortel is still pushing hybrids to protect their market and they were late to the pure IP game). The best pure IP solution based on price is ShorTel or Digium. But in the long run, these two have a higher risk of increased costs if they have to change their architecture or handsets or applications.

My final recommendation to future readers is this, if you don't have the expertise in house, then invest in the future of an employee, get him a lab and some training, sign him to a 2 year contract and but them in charge of finding a vendor. No one would put in a huge MS Exchange system without an MSCE in-house but apparently they fully expect to get a new IP system without having an IP Tel trained admin. This comes from the telecom industry sucking money out of users with MAC contracts and maintenance agreements.

__________________
Deus ex machina

fairways

I'm sorry that you are having problems with your integration/migration project. Can you tell me what features you are talking about (The Cisco sets will not do everything that Nortel will)?

Thanks,

Bill

telcoDave

aikenmarvin, I'm feeling your pain.

Got CCM 4.1 and I have to admit for the $$$, it is REALLY undernourished. A few new worthless features here and there but, hey, it's new (thanks, IT dept: ) Only quit twice in two years (even with redundancy). Got a Norstar with sets in critical departments. Worked like a charm and we survived.

In reply to fairways and our experience:
We used up a vacant dual PRI port and linked it up with a 2811 GW
Then we dealt with the misery of getting the 2811 to talk to anything other than itself. Was a couple years ago but I seem to recall that Cisco's idea of NI2 and Nortel's were different (not sure who won that war - we gave-up on it late one night, came back the next day and it was running! T1 elves I guess).
Then we set up CDP on the Nortel switch (have to look up the LD's) and began re-routing each DN (800 license and we were full) until we had nothing but trunks running through the Nortel.
Got new T1's and came in late with SBC, did the cut over, and powered down the ol' 61c (sniff, sniff).

All in all it was an experience. One that I can put on the resume. Llllottts of visits to PBXInfo.