Phone redunacy recommendations

[loneWolf9]

Currently we are working with a vendor that has sold us a solution, only to have THERE techs tell us it can not recover as quick as we were told.

Can anyone tell me what is considered standard recovery time between a prime, and backup site.
Looking for times that a PBX, App servers would roll from prime to backup equipment.

Need to cover the following
300 agents
advanced routing
screen pops
IVR.

Any thoughts would be appericated.
Thanks

 

[kwbMitel]

What you appear to be asking for doesn't exist. There is no "Standard Recovery Time"

The way a system is designed is fully dependant on the end users requirements and money available to make it happen.

Not all end users can afford to have fully redundant seamless disaster recovery sites. None exist in my experience. Everything is compromise between what the customer wants and what they can afford.

To answer the question with questions. How much time was disigned into your system for failover? Does it meet your design requirements? Do you have anything in writing?

*******************************************************
Occam's Razor - All things being equal, the simplest solution is the right one.

 

[loneWolf9]

We were sold a system in with redunat core, and app servers at to differenet locations. That if 1 system went down the other would be backup and running in a matter of mintues. As in 2-3min.

What we are being told now is the core (pbx) would take a min of 5 min. The app servers would at best case take 20 min.

We do have some in writing.

What I am looking for would be a real world response from real world techs.

What do you run (Avaya, Cisco, Nortel, Shortel)
What is your DR plan?
What is your recovery time.

A perfect solution does not exist. I know this.
However when promised a 3 min recovery, and then told 20 min is reality.

Can someone please give me some examples of solutions that they have in place?

 

[kwbMitel]

lonewolf9, you provide no specifics yet require specific answers. Life does not exist in a vacuum.

It sounds to me like someone either forgot to ask some key questions or some form of deception was intentional. An expectation of 2-3 minutes of downtime is quite exceptional and I would imagine, quite expensive.

As Ricky Ricardo would say "Someone's got some 'splainin to do"

P.S. I am a real tech in the real world.

*******************************************************
Occam's Razor - All things being equal, the simplest solution is the right one.

 

[loneWolf9]

KebMitel

I believe you to be a tech in real world. Never questioned anyone on this board about there creds.

Again I know what we have been promised, and the vendor.

As for specific's here you go. I would like to hear your reponse.

Requirements
Geo Redundant- with a DS3 dedicated connection.
Complete recovery 5 min or less.

Handle 300 agents initially, but grow to over 1000 agents at multiple locations

VOIP

Screen pops
Have to be able to work with our CRM software.
Real time reporting

Historical reporting

Advanced call routing, ACD

Ability to change agent’s skill sets on fly,
Automate roll over from skill set to another skillet at set periods of time in queue
Ability to have multiple skill sets for agents

Remote agents (ACD, and Standard)

Voicemail

Voicemail remote notification

IVR

 

[kwbMitel]

It all comes down to where the single points of failure reside.

With the product I am familiar with (Mitel, thus the name) all of your requests are achievable. With the possible exception of the skills changing on the fly depending on how that might be achieved.

I would design the system so that agents were load sharing between the Primary and Secondary sites. Under normal conditions all agents working without issues. In a failure, only half of the agents would need to re-sync. This would mean zero downtime for half the agents.

With infinite funding, all applications could be duplicated on redundant servers and would survive most failures as well.

With DS3 I would imagine you would be subject to the risk of someone cutting the fibre feed between buildings. Unless you have a redundant wireless solution as backup, you're hooped.

Unfortunately I can't speak for products other than Mitel as I've pigeon holed myself over the last 20 years.

*******************************************************
Occam's Razor - All things being equal, the simplest solution is the right one.

 

[loneWolf9]

Thank you for your feedback.

tell me have you ever migrated users over from Nortel to a Mitel solution.

 

[kwbMitel]

I've replaced Nortel systems with Mitel but none had ACD applications. I am not familiar with the Nortel offerings I'm sorry to say.

I'm surprised to hear you mention Nortel as it does not currently have the Voip growth potential you appear to need (as far as I know anyway). Maybe I've just been assuming Voip, my bad if so.

*******************************************************
Occam's Razor - All things being equal, the simplest solution is the right one.

 

[biglebowski]

Nortel could quite easily handle that growth. Both Barclays & the NHS call centres are currently running Nortel CS1000's which can handle 10's of thousands of users. It can also be networked to provide a single virtual call centre (NHS use this)

It's just not very advanced, reliable or easy to work with.

Sounds to me like they are pushing an Avaya solution which has quite good failover options but can be quite slow in handing over control to the backup CM's or AES's. I don't know about Aura which can be virtualised though as it's quite new.

Have a look at Siemens Openscape it provides a very good resiliency option but again is quite new so may be difficult to find impartial advice.

Did you put this out to tender, I would assume that you have already approched all the vendors for a solution/presentation if up time is of major importance to you.

You may be better off asking these questions in the specific forum for the vendor they are suggesting. There are Avaya, Nortel, Mitel, Cisco and Siemens forums on TT with some very knowledgable people.

"Anything that’s invented between when you’re fifteen and thirty-five is new and exciting and revolutionary, and you can probably get a career in it.
Anything invented after you’re thirty-five is against the natural order of things."

 

[kwbMitel]

Biglebowski, my comment regarding growth was for VOIP. Are you saying that Nortel now has a single controller platform that can handle 10's of thousands of VOIP phones. (Multiple controllers?) Again, I appoligise if my assumption regarding VOIP is groundless.


*******************************************************
Occam's Razor - All things being equal, the simplest solution is the right one.

 

[Sympology]

It also depends on your front end. You may want to pull some control away from the PBX's.

For example you could have Non Geo number with Instant fail over routes (we do this), that could route calls to an alternate PBX or Media Gateway (which itself has reduncy routing)then to the PBX.

Of course this means you would have to have a hot standby system constantly replcating data and the phones would also need auto failover (many SIP IP phones can do this).

the question is, do you want to do this?

1) You'd need PBX's capable of this
2)You'd need the infrastructure to do this (extra hardware)
3) And this is a overlooked one, do you want to automatically fail over or a manual process. Auto is good and can be fast, but could it be more hassle than it's worth? For example do you have auto revert or revert at a scheduled time. If you have a relativly small issue with a trunk, do you want your backups kicking in and potentially causing other issues?


Robert Wilensky:
We've all heard that a million monkeys banging on a million typewriters will eventually reproduce the entire works of Shakespeare. Now, thanks to the Internet, we know this is not true.

 

[biglebowski]

A single call server can handle up to 22,500 IP phones.

"Anything that’s invented between when you’re fifteen and thirty-five is new and exciting and revolutionary, and you can probably get a career in it.
Anything invented after you’re thirty-five is against the natural order of things."

 

[kwbMitel]

Thanks Biglebowski, as I indicated earlier my knowledge of Nortel is weak.

Knowledge is power.

which reminds me of a joke.
Knowledge is Power
Time is Money
and as every engineer knows, Power = Work over Time.

Substitute to get
Knowledge = Work over Money
The More You Know, the More Work You Do, and
The More You Know, the Less Money You Make.

or

Money equals Work Over Knowledge.
The More you Make, the Less you Know.

or

Work equals Money times Knowledge
As Knowledge apporaches zero, Work approaches Zero.

*******************************************************
Occam's Razor - All things being equal, the simplest solution is the right one.

 

[loneWolf9]

I have a co-worker with this posted at work.

We are now engaged with the design team of Avaya/Nortel.

I do not wish for this to become a bash post, so I will hold off on anymore comment until we have a final decision.

Thanks to everyone for their feedback.

I will post when all is said and done.