[BBAdmin]

Having a strange issue on one of my 4.1 BES servers (Exchange 2003 based). Over the last week I've been getting messages from my BES Alert system. I'll usually receive about 6 messages all at the same time. Example as follows:

Example 1

BlackBerry Dispatcher "servername" (Application Event Log on "servername"): 05/03/2006 10:01:19 (AFFF4E20) -> [SRP] Connection failed

Example 2

BlackBerry Dispatcher "servername" (Application Event Log on "servername"): 05/02/2006 00:41:44 (6FFFC3B0) -> [SRP] Dispatcher\SRP Connection established

Example 3

BlackBerry Dispatcher "servername" (Application Event Log on "servername"): 05/02/2006 04:43:10 (6FFFC3B1) -> [SRP] Dispatcher\SRP Connection dropped, Error=0


It looks as though for some strange reason the SRP connection drops and then reconnects itself almost instantly - really strange

Also had this one at the same time as the last cluster of emails from BES Alert:

BlackBerry Dispatcher "servername" (Application Event Log on "servername"): 05/03/2006 10:01:01 (AFFF4E20) -> SRPClient::Connect: Error calling host "localhost" [127.0.0.1] (10054)

I can't seem to find anything wrong and because it all happens so quickly it's not like any users have noticed anything. I have not dived in as far as the log files just yet (avoiding those at all cost)!!!

Anybody come across anything like this????

[tickler]

I'm running BES Version: 4.0.1.11 and I've seen this error recently in the logs

SRPClient::Connect: Error calling host "localhost" [127.0.0.1] (10061)

Not sure how long it's been happening but I noticed it because I'm looking into a problem with pending mail on the BES, some users are not getting mail on the device, after a server restart all the mail is delivered.

Originally Posted by BBAdmin

Having a strange issue on one of my 4.1 BES servers (Exchange 2003 based). Over the last week I've been getting messages from my BES Alert system. I'll usually receive about 6 messages all at the same time. Example as follows:

Example 1

BlackBerry Dispatcher "servername" (Application Event Log on "servername"): 05/03/2006 10:01:19 (AFFF4E20) -> [SRP] Connection failed

Example 2

BlackBerry Dispatcher "servername" (Application Event Log on "servername"): 05/02/2006 00:41:44 (6FFFC3B0) -> [SRP] Dispatcher\SRP Connection established

Example 3

BlackBerry Dispatcher "servername" (Application Event Log on "servername"): 05/02/2006 04:43:10 (6FFFC3B1) -> [SRP] Dispatcher\SRP Connection dropped, Error=0


It looks as though for some strange reason the SRP connection drops and then reconnects itself almost instantly - really strange

Also had this one at the same time as the last cluster of emails from BES Alert:

BlackBerry Dispatcher "servername" (Application Event Log on "servername"): 05/03/2006 10:01:01 (AFFF4E20) -> SRPClient::Connect: Error calling host "localhost" [127.0.0.1] (10054)

I can't seem to find anything wrong and because it all happens so quickly it's not like any users have noticed anything. I have not dived in as far as the log files just yet (avoiding those at all cost)!!!

Anybody come across anything like this????

[BlackBerryLinks]

Some times it is because Windows Policies where changed and or other software is installed on the BES.
Check the services:
http://blackberry-nl.com/index.php/topic,205.0.html

It is possible that there is a TCP/IP port conflict with other software (e.g. Backup software).

Regards,
Bob

[bbgirl]

We just upgraded to 4.1 and we get it too.. My CIO wants to know what the deal is...

There isn't anything else installed on my BES. Wouldn't I get an error regarding an tcp/ip conflict?

[bbgirl]

[RELAY_SRP:UNKNOWN:013faf68] Connecting to relay at address "srp.na.blackberry.net"
[30000] (07/31 15:31:16):{0x524} [RELAY_SRP:UNKNOWN:013faf68] Connect: Found 2 addresses for host "srp.na.blackberry.net"
[30000] (07/31 15:31:16):{0x524} [RELAY_SRP:UNKNOWN:013faf68] Connect: Host "srp.na.blackberry.net" @ [xxx.xx.x.xxa]
[30000] (07/31 15:31:16):{0x524} [RELAY_SRP:UNKNOWN:013faf68] Connect: Host "srp.na.blackberry.net" @ [xxx.xx.xxb]
[30000] (07/31 15:31:16):{0x524} [RELAY_SRP:UNKNOWN:013faf68] Connect: Trying "srp.na.blackberry.net" @ [xxx.xxx.xx.xxa]
[30000] (07/31 15:31:16):{0x524} [SERVICE_RELAY_SESSION:UNKNOWN:013f73d8] Connection to relay established.


okay, I read this in my router log.. at the same time my SRP drops...

anyone have any ideas? I have one, but would like to hear others.

[waddyjohn]

Its probably the dispatcher crash related bug that was supposedly fixed in SP5 for 4.0. Its supposed to be resolved with SP1 for 4.1.

[bbgirl]

something else I noticed..


has there always been two hosts for each..

blackberry nodes.


I noticed that my srp makes two connection sessions to a different srp host.


anyone else have this... both are to srp.na.blackberry.net, but with two ip's.

[|||||||]

Originally Posted by bbgirl

anyone else have this... both are to srp.na.blackberry.net, but with two ip's.

Yes if you do an nslookup of srp.na.blackberry.net you will see two IPs, this is for redundancy.

[bbgirl]

I thought that, but only thing is my srp drops for about 2 sec now...and I looked in the logs, and noticed that when it drops its doing that connecting to the relays..

[dev]

Check your ethernet connection to the network devices ( switch, routers, firewall). Also check the firewall logs for the "connection state" of the BES to the SRP host. You may find that the connection was dropped at firewall level.
One of the most common network performance issues is related to ethernet duplex mismatch. This occurs when one port of the link is operating at half-duplex (send or receive) while the other port is operating at full-duplex (send and receive simultaneously).
Some routers re-negotiate auto-speed or auto-duplex which can introduce jitter (latency). Modern devices can auto-sense speed and duplex, while some older devices can only operate in half-duplex mode.
This latency (jitter) can cause the SRP connection to be dropped and re-established again.

[bbgirl]

Originally Posted by dev

Check your ethernet connection to the network devices ( switch, routers, firewall). Also check the firewall logs for the "connection state" of the BES to the SRP host. You may find that the connection was dropped at firewall level.
One of the most common network performance issues is related to ethernet duplex mismatch. This occurs when one port of the link is operating at half-duplex (send or receive) while the other port is operating at full-duplex (send and receive simultaneously).
Some routers re-negotiate auto-speed or auto-duplex which can introduce jitter (latency). Modern devices can auto-sense speed and duplex, while some older devices can only operate in half-duplex mode.
This latency (jitter) can cause the SRP connection to be dropped and re-established again.

I have been trying to chat with our firewall woman, today. But she has been busy. I was told that a new switch was put in place, but apparently... some things have changed or something.

[andidarmali]

Example 1

BlackBerry Dispatcher "servername" (Application Event Log on "servername"): 05/03/2006 10:01:19 (AFFF4E20) -> [SRP] Connection failed

Example 2

BlackBerry Dispatcher "servername" (Application Event Log on "servername"): 05/02/2006 00:41:44 (6FFFC3B0) -> [SRP] Dispatcher\SRP Connection established

Example 3

BlackBerry Dispatcher "servername" (Application Event Log on "servername"): 05/02/2006 04:43:10 (6FFFC3B1) -> [SRP] Dispatcher\SRP Connection dropped, Error=0

I've ever experienced something like that, and it happened when the GPRS connection is under maintenanced. Maybe you should ask your provider whether they're doing some maintenance or not?

[Crowdie]

I had a similar problem at a customer's site running BES for Domino 4.0.1. At around 23:15 the BES kept losing connection with the RIM router until around 23:45. The fix was to install BES SP4 as this corrects an issue where the BES loses connection with the RIM router while doing maintainence on the configuration database (normally at night).

[firedragon]

srp connection lost, this happens from time to time. If you review your log (app log and dispatcher log), you will see that. This is OK as long as you see srp connection establised msg in your log. I think rim code pings it's srp conenction to keep it alive...

srp connection failed is when your srp id has been disabled by RIM. Verify this on the BlackBerry Manager. If your icon has a red X on it, that's the same indicator. If it is the case, call RIM and ask them to reenable to srp id for you. Typical scenario is when BES Dispatcher tries to reconnect more than 5 times in 1 minute, the srp id will get disabled by RIM. This is a RIM rule in their code by default. There is a workaround for it. Go to Rim pubic knowledge center and take a look at KB-05278. By increasing the reconnect time interval to 15 seconds will fix this problem.

Hope this helps.

[bbgirl]

Originally Posted by dev

Check your ethernet connection to the network devices ( switch, routers, firewall). Also check the firewall logs for the "connection state" of the BES to the SRP host. You may find that the connection was dropped at firewall level.
One of the most common network performance issues is related to ethernet duplex mismatch. This occurs when one port of the link is operating at half-duplex (send or receive) while the other port is operating at full-duplex (send and receive simultaneously).
Some routers re-negotiate auto-speed or auto-duplex which can introduce jitter (latency). Modern devices can auto-sense speed and duplex, while some older devices can only operate in half-duplex mode.
This latency (jitter) can cause the SRP connection to be dropped and re-established again.

I should probably ask them to check at the time of lost connectivity huh? but I like that answer about full and half...we have to change our desktop cards all the time, because this place is so screwed up. thanks.