BlackBerry Forums Support Community               

Closed Thread
 
LinkBack Thread Tools
Old 08-07-2006, 11:52 AM   #1 (permalink)
New Member
 
Join Date: Jun 2005
Model: Curve
Carrier: AT&T/Cingular
Posts: 9
Post Thanks: 0
Thanked 0 Times in 0 Posts
Exclamation SRP Disconnect/Disabled Issue

Please Login to Remove!

Recently we have developed a problem with SRP connection drops leading to the 5-in-1 rule that disables our SRP ID.

We have been using Blackberry for over three years now. Until May we have never had our SRP disabled due to the 5-in-1 rule (5 reconnects within 1 minute). Prior to this we were stable on Domino BES 4.0 SP2 HF2 for six months. On 7/16/06 we updated to SP4 HF3 with the template hotfix, and the issue was not affected.

Our versions are Domino 6.5.3, BES 4.0.4.9, Windows Server 2003 SP1, running on HP DL380s. We have two production BES SRPs, and each one has had the problem. Sometimes the disconnects occur on both servers at the same time. At other times there is no coorelation between the two production server disconnects. We have averaged one disconnect every two weeks since May with multiple disconnects on the same day on some occasions.

We have reviewed our internal network and firewall configurations and found no issues. We upgraded our internet connection and swapped out the connection hardware.

We added the Dispatcher registry values from support to prevent the 5-in-1 rule by lengthening the reconnect interval, but that has not helped. We confirmed with RegMon that the Dispatcher service picks up the registry values, but it does not use them.

We are seeking feedback from other customers who have struggled with SRP disconnects recently. What did you find the root cause to be? Please share specifics if you can.

Thank you.
Offline  
Old 09-19-2006, 04:19 PM   #2 (permalink)
Knows Where the Search Button Is
 
PlatzDa's Avatar
 
Join Date: Mar 2006
Location: Sacramento, CA
Model: 9630
OS: 5.0.0.975
Carrier: Verizon
Posts: 20
Post Thanks: 0
Thanked 0 Times in 0 Posts
Default

I was just now searching for info on this issue. We lost our connection on Friday, and our logs (server and firewall) revealed no problems on our end. We got it reconnected by RIM, but the tech was unable to tell me how it happened. This leads me to believe that the problem was either caused by something here that we haven't found yet, or RIM is blowing smoke and not revealing that they screwed up somehow (although I can't imagine how).

Like many of you, I became a BES admin through no choice of my own (and without any training, of course), so forgive me if this all sounds unintelligent...
__________________
"When the going gets weird, the weird turn pro"
Offline  
Old 09-19-2006, 04:47 PM   #3 (permalink)
New Member
 
Join Date: Jun 2005
Model: Curve
Carrier: AT&T/Cingular
Posts: 9
Post Thanks: 0
Thanked 0 Times in 0 Posts
Smile SRP issues fixed in 4.0 SP5 HF2

Our SRP issues were specifically addressed in 4.0 SP5 HF2. We have been running it on both of our production servers now for two weeks without issue. The SRP disconnects have disappeared.
Offline  
Old 09-19-2006, 09:43 PM   #4 (permalink)
Talking BlackBerry Encyclopedia
 
dropscience's Avatar
 
Join Date: Jul 2006
Model: 8800
Carrier: Rogers
Posts: 216
Post Thanks: 0
Thanked 0 Times in 0 Posts
Default

Let me share my experience this morning...

Our BES server wasn't operating and I noticed the SRP status was disconnected. Port was still 3101 and it was pointing correctly to http://srp.na.blackberry.net but the SRP test connection test on the manager failed.

Like you guys, nothing in the event log identified this disconnect so we traced the DNS resolution of srp.na.blackberry.net and everything was resolving "correctly" according to the firewall so why aren't we connecting to RIM?

We later found DNS corruption on our Active Directory side. Somehow it was resolving to some place in New Orleans??? DNS resolution of the SRP was incorrect in our AD cache! we dumped the cache and rebooted the BES - bingo, it came up!

That's my story... our Network admin called it DNS poisoning but where/how/why it happened who knows??
Offline  
Old 09-20-2006, 09:58 AM   #5 (permalink)
BlackBerry Extraordinaire
 
noname's Avatar
 
Join Date: Sep 2005
Location: Congested Islet of "Foreign Talents" (> 45% of workforce) - Singapore.
Model: Z10
OS: 10.0.0
PIN: NUKE(PAP)
Carrier: Singtel
Posts: 1,504
Post Thanks: 6
Thanked 9 Times in 9 Posts
Default

There is a known issue in SP4 that SRP will be disabled if the dispatcher is sending a large packet to router due to a large attachment processing. If you search the router log, you will see entries with "Too_big". If this is the case, upgrade to the latest SP5 HF#2 and it should resolve the issue. Otherwise, you will need to perform the workaround i.e. to shut down the server, get RIM online to re-enable your SRP and then restart the server. Hope this helps!
__________________
Native but 4th class citizen of a nation governed by idiots who import congestions & contention.
Offline  
Old 09-20-2006, 10:06 AM   #6 (permalink)
CrackBerry Addict
 
|||||||'s Avatar
 
Join Date: Jun 2006
Model: 7100
Carrier: Rogers
Posts: 615
Post Thanks: 0
Thanked 0 Times in 0 Posts
Default

You don't need to shut down your server to have the SRP re-enabled.
Offline  
Old 09-20-2006, 10:13 AM   #7 (permalink)
Knows Where the Search Button Is
 
Join Date: May 2005
Model: 8800
Carrier: O2
Posts: 36
Post Thanks: 0
Thanked 0 Times in 0 Posts
Default

Before you get RIM to reactivate the SRP run these steps.

1) Validate the SRP/Authentication Key in the Blackberry Server Configuration. If the key is locked it will return an Invalid Key response. be patient has this can take a minute or two to reply if locked and it will look like the application as hung while thinking about it. If it is locked, it will return an invalid key message.

2) Disable Powerpoint viewing in the attachment server configuration, this is the issue which is causing the data "Too_Big" issue in the router logs.

3) Get RIM or your service provider to unlock the SRP key. If you still have the powerpoint attachment viewing enabled, you can lockout your SRP key again.

4)Then reboot your server.

5) Plan to update your BES install to SP5a as this resolves the powerpoint attachment issue whic causes SRP lockouts.
Offline  
Old 10-04-2006, 10:48 AM   #8 (permalink)
Knows Where the Search Button Is
 
crstek's Avatar
 
Join Date: May 2006
Location: Chicago, IL
Model: 7520
Posts: 44
Post Thanks: 0
Thanked 0 Times in 0 Posts
Default

Our box just went down today. I flushed dns/rebooted the box twice, no avail. I also tested the SRP connection, which is fine. We are on hold with RIM right now...any suggestions...
Offline  
Old 10-21-2006, 11:34 AM   #9 (permalink)
Talking BlackBerry Encyclopedia
 
Mark_Venture's Avatar
 
Join Date: Nov 2005
Location: Delaware
Model: 8900
Carrier: T-Mobile (w) - Verizon (P)
Posts: 313
Post Thanks: 0
Thanked 0 Times in 0 Posts
Default

Ran into this too... We have two BES's... for this discussion we'll call them Bes1 and Bes2...

We ran into this issue of "packet size greater than 64000" being sent to RIM's network 5 times in under 1 min... (Yeah, fixed in service pack)... which caused RIM to disable our SRP key for BES2 that this happend to... BES1 was up and functioning normally...

The only "error" condition we saw was the red X on the BES name, and saying Disabled in the blackberry manager software.

The SRP Connection test button on the Blackberry Router tab of the Blackberry Server Configuration program worked...

Of course it needed a call to RIM, and having them re-enable the SRP key, then explain this error to me, and help me find it in the error logs...

So now the server team and operations are trying to add netiq monitoring to my server now, and want to monitor that SRP connection...

Applying Service pack 5A and hot fixes later today to both servers to help prevent in the future...

But now I have to try and SIMULATE this failure so they can test the script... but I have no clue how.
Offline  
Old 11-01-2006, 06:26 PM   #10 (permalink)
Knows Where the Search Button Is
 
PlatzDa's Avatar
 
Join Date: Mar 2006
Location: Sacramento, CA
Model: 9630
OS: 5.0.0.975
Carrier: Verizon
Posts: 20
Post Thanks: 0
Thanked 0 Times in 0 Posts
Default

The fix for our problem was to change the registry value that disables the SRP connection after 5 attempts. We seem to have this problem due to the super-fine quality of our network infrastructure...

The reconnect timeout can be changed to 15 seconds, which will mean that only 4 reconnect attempts can be made per minute, which can help you avoid having to contact RIM to get you reconnected! See RIM KB article KB-05278 for more info...
__________________
"When the going gets weird, the weird turn pro"
Offline  
Closed Thread


Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On





Copyright 2004-2014 BlackBerryForums.com.
The names RIM and BlackBerry are registered Trademarks of BlackBerry Inc.