powered by Jive Software

Still experiencing random disconnects in Spark

I was seeing this around 2.5.7 (and possibly 2.5.6). I have upgraded the server to 3.4.2 (I think I recall seeing this as early as 3.3.2). I’ll be in a conversation with someone and then for no apparent reason I am disconnected. I wait 10 seconds, and I get reconnected. What is going on?

Is nobody else experiencing this??

Ok, this is REALLY getting bad now.

I’m running Spark 2.5.8 and OpenFire 3.4.4, Java Runtime 1.6_0_5. And I am getting disconnected during almost ANYTHING I do.

If I send a text file to someone, I get disconected.

If I’m typing to someone, I get disconnected.

WHAT is going on? What can I provide to help diagnose this problem? The software used to be rock solid, but this is getting ridiculous for me.

I havent seen such issue. Maybe there is a problem in connection between your Spark and Openfire server? I mean your network.

— I stripped out the logs entries & posted a more complete log later in this post. HELP ME!

I am having the same problem. Please reply or email me if there is a known bug and/or bug fix for this.

Wayne DeBee

The InventSAI Network - wdebee@inventsai.com

Update: The problem I am having appears to be isolated to Windows Vista machines. It is constantly disconnecting and reconnecting. These two machines are for outside sales employees connecting through the internet. All machines on my internal network are running Windows XP. None of the XP machines are experiencing this problem. I am no expert, but after reviewing this a little I came across some information on User Account Control (UAC) for Windows Vista. Here is the link.

http://technet.microsoft.com/en-us/windowsvista/aa905108.aspx

I have not followed up on this because I don’t have a Windows Vista machine to research it. If anyone knows anything about this, or can figure out a fix please let me know so that I can forward it to my outside people.

Thank you.

Does the fact that I’ve got dual NICs and Windows network connection sharing have any bearing on this?

Ok, it’s getting worse & I’m starting to get client complaints. I’ve attached the logs.

If this keeps up, I’m going to have to drop Spark as my default client, which I really don’t want to do. But I can’t keep having the messenger service randomly goingh up & down like kangaroos in the mating season.

HELP!!!

Also, here are the logs from the actual OpenFire server itself.

Hi,

“org.apache.mina.filter.codec.ProtocolDecoderException: java.lang.StringIndexOutOfBoundsException: String index out of range: 0 (Hexdump: empty)” looks like a MINA problem. This should not happen when sending ASCII characters to JIDs with ASCII characters but as soon as you have some non-ASCII characters involved you could get this error. See Problem with cyrillic in messages / JM-1175 for a solution (downgrade MINA).

Disable compression if it is enabled as this can also cause problems.

There are also some DB connection issues in your log file, but they are likely not related to disconnects. Openfire 3.4.5 should solve this problem with the new connection pool.

LG

Hi,

it seems that also the IM Gateway update did fail for you. You may need to update the database manually.

LG

Well I have taken the postings to date into consideration, done some clean up, and am still experiencing issues. Though now, I am unsure what is causing the problems. Here is what I have done:

First, my Openfire server is running inside a Virtual Machine. On the VM host, I have 6 virtual machines in total: 3 web, Openfire, mySQL and MS SQL.

I noticed that the traffic times were very erratic going from the mySQL server to the Openfire server. What I did was to give both machines (Openfire & mySQL) a second network card to a VM switch so the two machines are now networking at gigabit/memory speeds. When I do a ping test, the pings are now consistantly <1ms wherea before they could reach up to 400ms.

I rebooted the server and all seemed well. About an hour later, everyone disconnected again.

I then recalled the comment about the misconfigured/update plug in. I stopped the server, removed every unnecesary/unconfigured plug in (and the corresponding folders) and restarted the server. Shortly thereafter, another crash (the only plug ins in my folder are: broadcast, motd, search, sparkmanager, and userImportExport).

I also noticed that my Spark logs and Openfire logs were filling up as a result of the disconnects. To aid in debugging, I have flushed all the logs. I have also, for the time being turned on debugging. I’m not sure if this plays into it or not, but this is what was in the debug screen (see attached text file):

I am currently running:

Openfire 3.4.5

MySQL 5.x

Java Runtime Environment 1.6.0_04_b12

Windows XP SP2

The clients should all be running Spark 2.5.8 with a minimum of JRE 1.6.

If I receive another crash/disconnect I will post the shortened logs to aid in debugging. Any ideas?

Ok, I just experienced my first disconnect since I made the previous post.

I checked the error log on the server, and nothing. Also, nothing in warning or info.

However, the debug area showed something (see disconnect-spark-log.txt) and I’m attaching the log from Spark on my machine.

Any idea why this would be happening?

I had two users switch their clients from Spark to something else (Pandion and Psi). The Psi user was disconnected as well, so I’m not sure what the issue is.

Would any invalid database entries cause a problem? I had to do a wonky DB update/migration a ways back. I’m wondering if it’s now manifesting itself. I could post/send a copy of the database if that would help.

I am having the same issues on Openfire 3.6.4 and Spark 2.5.8 with AD and emded DB- random disconnects to random users. Rebooted Openfire server and domain controllers. All settings look fine.

Any help?