RIMs Main Servers

[rcabls]

I'm new to this and was curious as to where the main servers are located. I understand RIM has servers in Canada and Great Britain, but how many exactly? I know all data traffic passes through one of these servers, but what are they called and is that their only function? What about Asian customers? Do they pass through these servers as well? I have been trying to find out more information on these servers but can't seem to come up with anything informative. Can anyone here help me out?

[CanuckBB]

All BB data goes through RIM's infrastructure. The main datacentre is believed to be located in RIM's headquarters in Waterloo, Ontario, Canada.

Any other information is a very closely guarded secret for obvious reasons.

Why are you asking?

[BBAdmin]

Closely guarded..........there's one in North America, one is Europe (which is in the UK) and one in Asia Pacific. I believe they are now breaking these three down to cover sub territories as well. For example, I believe there are now two relays that cover Europe. Can't be that closely guarded, I've had a tour of the UK one!!!!

[rcabls]

I am curious because as a customer I want to have an idea of how my information is routed and who has access to it. In the US my provider controls the traffic and I wouldn't want a competitor to have access to the data. Can I be sure to trust Canada? More important, after their network went down a year or so ago, I'm curious as to how they can guarantee service with what seems like a bottle-neck infrastructure. If the Canada servers go down again as they did a year or so ago, what is their backup plan?

When I went through some of the documentation from the Blackberry site, it really doesn't mention their servers in the topology. They just show pictures of the handheld to the wireless network and then to your firewall and network. They neglect to mention the RIM servers that are placed along the way.

[JSanders]

I wouldn't want a competitor to have access to the data. Can I be sure to trust Canada?

What would they do with it? Make more beer out of it?
j/k

But seriously, with all the data that goes through their servers, you really your information would be a target, EVEN IF your competitor "could" gain access to it? If that is the case, our US national security is poopie.

[BBAdmin]

It's a silly question. I don't think they can sniff the packets to see what's in them anyway - it's just an encrypted packet of ones and zeros, no spying involved!!! This is a FIPS 140 model remember.

[rcabls]

True, there isn't much of a threat. Although I have been suspicious of our Canadian neighbors lately...they're up to something! I'm sure its no big deal, but I was researching into this since I made my original post and it seems the Waterloo NOC may have had a fault when the outage in August happened. My question is do they have another reliable NOC in place? I also wonder if data could default by going through the UK NOC instead? It seems unlikely, but it would make me feel safer knowing there was a backup in place.

[JSanders]

I know, ever since they lowered the limit on the daily bag on offshore salmon, I have been keeping an eye on them. They COULD be trying to protect all those icebergs from global warming, but do they care?

I don't recall an outage in August. I do recall one in April.

[vliou]

You're doubting trusting Canada? A Neutral country? You're mad, it's the Americans I never trust with...killing each other, killing other countries, installing back doors in Windows...

Come to Canada. We don't treat you like terrorists here.

[JSanders]

lol, but I was only joking, you seem frightfully serious.

[vliou]

*Gasp*

I was *SHOCKED*, *SHOCKED* to see Canada listed as a state in the Virginia DMV.

Your US dollar has been ticking me off, as well as the CDN dollar being unrealistically high. Any "pissedoffness" can be attributed to our current financial market!

[JSanders]

lol, yea.

by the way, you are needed over here in this thread, badly. SW Virginia, right?
http://www.blackberryforums.com/rants-raves-forum/94709-rant-morals-values-honesty-2.html#post694206

[rcabls]

Back to the topic, the only confirmed RIM NOC's are in England and Canada? It was suggested there was one in Asia-Pacific, but from my research I can't confirm this.

[loganki]

I think its a fair and interesting question. I frequently ask the same question to the admins of networks I use (ie. my schools email servers)

[rjw3000]

Have you emailed RIM your question? Worst that could happen is they don't respond or say no comment.

[juwaack68]

Back to the topic, the only confirmed RIM NOC's are in England and Canada? It was suggested there was one in Asia-Pacific, but from my research I can't confirm this.

Don't type so loud. "They" can detect loud packets and will zero in on them.

Especially those Canadian Loonies.

[JSanders]

Back to the topic, the only confirmed RIM NOC's are in England and Canada? It was suggested there was one in Asia-Pacific, but from my research I can't confirm this.

Then that makes it not true?

[juwaack68]

Then that makes it not true?

I think I read about it on my Bazooka gum wrapper. But that was in code and I lost my super decryption ring from the Fruity Pebbles box.

[rjw3000]

From reading a bunch of the OP's questions, does he seriously think specific info about servers and how everything works is going to be readily available? I can see them giving general info like where they have NOC's and stuff, but a bunch of the other questions, come on!

As far as things going down and backup plans, most big companies have enough servers and redundancy to where it might appear to the end user that the whole system is down, but in reality only some servers are down and they are hitting those servers that are down.

[hdawg]

$0, J00 w4nt To kNOW @<hidden> 48oUt r1M @<hidden> tHEiR N0c$ ANd H0w TH3Y WoRK? ... s33 tHE lINK$ B3L0W. +H@<hidden>'$ nOT TO 5ay +h4+ THEY DoN'+ H@<hidden> 0+H3r nEtW0rk5 1n +H3 M1DdL3 +h4T rOUt3 TrAfF1c ... perh4P$ Thr0uGh @<hidden>@<hidden>+4n.

Firewall and connection requirements for the BlackBerry Enterprise Server (http://www.blackberry.com/btsc/search.do?cmd=displayKC&docType=kc&externalId=KB03735&sliceId=SAL_Public&dialogID=56110903&stateId=0%200%2030567428) and Firewall and connection requirements for the BlackBerry Internet Service (http://www.blackberry.com/btsc/search.do?cmd=displayKC&docType=kc&externalId=KB11036&sliceId=SAL_Public&dialogID=56110903&stateId=0%200%2030567428)

... If you don't trust that the data being passed in / out of your network is secure, then either:

1) Don't use the service
or
2) Capture all the traffic and try to decrypt it.

haXoR +He pl4NE+!!!!

[rcabls]

To clarify what I would like to know, I am not looking for specifics into how many servers they are running or what not. I was wondering about the single point of failure issue that brought the network down a while back. I imagine RIM has server farms set up in Ontario, England and Asia Pacific, but I can't find official documentation to verify this.

Why does this matter? For instance, if the Ontario servers go down does the traffic route through England or Asia Pacific? If not, what is the plan for fault tolerance? Is there a guaranteed time frame for when my users can send and receive messages on their devices?

This isn't about the network specifics, I am looking at a higher level and it's also worth pointing out that if one of the other server regions goes out, will that route all of that traffic through the other servers as well? It might make more sense to stay with a cellular provider who does not have a central point of failure in this case.

[hdawg]

If an entire region were to go down RIM would have to make a transparent change for failover to occur as SRPs are tied to specific NOCs ... they connect to a NOC and that is their NOC; you can just change from srp.na to srp.eu

If an entire srp network goes down, I know of no plan. example:

My company has 2 data centers ... if both go down, the plan is to go to our tape backup location and restore the network at the presidents house (or my house or something) ... The fault tolerance is that we have 2 data centers; if both die that isn't fault tolerance but rather disaster recovery. RIM hasn't had a disaster yet ... but you raise an interesting point about what happens if an entire SRP network goes down.

I guess I kind of accept the fact that they'll be there; much like the root servers or my ISPs mail servers that I route through.

[rsnadel]

To clarify what I would like to know, I am not looking for specifics into how many servers they are running or what not. I was wondering about the single point of failure issue that brought the network down a while back. I imagine RIM has server farms set up in Ontario, England and Asia Pacific, but I can't find official documentation to verify this.

Why does this matter? For instance, if the Ontario servers go down does the traffic route through England or Asia Pacific? If not, what is the plan for fault tolerance? Is there a guaranteed time frame for when my users can send and receive messages on their devices?

This isn't about the network specifics, I am looking at a higher level and it's also worth pointing out that if one of the other server regions goes out, will that route all of that traffic through the other servers as well? It might make more sense to stay with a cellular provider who does not have a central point of failure in this case.


My advice would be to find another provider. Clearly you have no faith in RIM and you seem to have a paranoia complex that rivals anything I've seen. If the security issue is that big a deal for you, I don't see how anything short of you actually rebuilding the system yourself will satisfy you. The executive branch of the U.S. government uses BB, but you find them woefully short in data integrity protection. Go figure.

[BBAdmin]

You do seem to be more paranoid than a bag or rats in a burning meth lab.

[CanuckBB]

A balance needs to be struck between fault tolerance and costs. Personnaly, I don't want to pay more to support the kind of infrastructure that will avoid a down network overy few years. I've been in I.T. for far too long. I do understand 'Shit Happens'.

If you think that access to email is that critical to YOU, YOU need a contingency plan.

[joginder]

i donno how much info we have got hre thru this thread because so far everything is hypothetical and ideas or guess work. I dtrongly believe that RIM's NOCs are operating with fail safe and fail over and would route the traffic thru other relays. the only issue that I would see if they x-fer the info thru other relays like all the BB data for fed govts starts flowing thru Asia/China and that will be a concern for NSA or others and thats why i would see RIM not routing the messages thru other networks.
just my 0.002$

[Keyscan]

This is one question you will not get full answers to, I can pretty much promise you that.

[Keyscan]

A balance needs to be struck between fault tolerance and costs. Personnaly, I don't want to pay more to support the kind of infrastructure that will avoid a down network overy few years. I've been in I.T. for far too long. I do understand 'Shit Happens'.

If you think that access to email is that critical to YOU, YOU need a contingency plan.

RIM goes down twice a year for 6 hours. The carriers go down all the time. Most times when people think the BIS is down, it is actually their cell phone providers connection to the BlackBerry Infrastructure that takes a nose dive.

[Berry One]

It might make more sense to stay with a cellular provider who does not have a central point of failure in this case.

Name one.

[CanuckBB]

RIM goes down twice a year for 6 hours. The carriers go down all the time. Most times when people think the BIS is down, it is actually their cell phone providers connection to the BlackBerry Infrastructure that takes a nose dive.

The BIS infrastructure may g down more often. The BES always never does. Certainly not twice a year for as long as 6 hours. Other than the one last year or do that lasted 12 hours, I can't recall the BES infrastructure going down long enough to notice.

[Keyscan]

The BIS infrastructure may g down more often. The BES always never does. Certainly not twice a year for as long as 6 hours. Other than the one last year or do that lasted 12 hours, I can't recall the BES infrastructure going down long enough to notice.

You are right, the BIS technically does go down more often by the looks of it. But entire 100% outage of the BIS really only happens a few times a year. Almost every time I see it is because of a carrier outage.

When the "BES" so to speak goes down, that means everything is down, including BIS. Everything needs to run through the "BlackBerry Infrastructure" and that is what went down. You are right though, that hardly ever goes down at all and thank god for that... :)