(EventWriter.processEvent): failed writing to channel

rewb0rn · Post by **rewb0rn** » 26 Jan 2009, 12:57

Hi,

we have several log entries like this one:
2009/01/21 17:29:11.057 - [ FINE ] [id: 158] (EventWriter.processEvent): failed writing to channel: java.nio.channels.SocketChannel[connected local=/0.0.0.0:0 remote=/91.7.28.176:2602]
2009/01/21 17:29:11.059 - [ FINE ] [id: 12] (SmartFoxServer.lostConnection): User [ JackPott_64463 ] removed

Can someone give me a hint why they are generated and what can be done to avoid user disconnection if there is any way?

Thanks in advance.

Post by **Lapo** » 26 Jan 2009, 16:17

What server version is it?

rewb0rn · Post by **rewb0rn** » 28 Jan 2009, 11:14

Our customer is using 1.6.4 with which these messages were generated but I am developing with 1.6.3 locally if that makes any difference. I just did notice that there is actually a version difference, I will update asap.

whornak · Post by **whornak** » 01 Feb 2011, 21:47

Was there a resolution found for this, I am getting this error message consistently with one tester.

OpenSpace 2
SmartFoxServer 1.6.9

2011/02/01 17:45:29.304 - [ INFO ] [id: 14] (Logger.logInfo): [OpenSpace] Map load request received from user MarcButton [17]
2011/02/01 17:45:29.304 - [ INFO ] [id: 14] (Logger.logInfo): [OpenSpace] Joining user MarcButton [17] to Room 'E2' [40]
2011/02/01 17:45:29.304 - [ FINE ] [id: 14] (User.deleteVariable): User variable [_os_pz] deleted for user: MarcButton
2011/02/01 17:45:29.304 - [ FINE ] [id: 14] (User.deleteVariable): User variable [_os_px] deleted for user: MarcButton
2011/02/01 17:45:29.304 - [ FINE ] [id: 14] (User.deleteVariable): User variable [_os_py] deleted for user: MarcButton
2011/02/01 17:45:29.304 - [ INFO ] [id: 14] (Logger.logInfo): [OpenSpace] Room 'E2' [40] joined; now loading associated map 'bpb_E2#E2_Map
2011/02/01 17:45:29.304 - [ INFO ] [id: 14] (Logger.logInfo): [OpenSpace] Loading assets library for map 'bpb_E2#E2_Map'
2011/02/01 17:45:29.304 - [ INFO ] [id: 14] (Logger.logInfo): [OpenSpace] Library 'bpb_E2' already available in cache
2011/02/01 17:45:29.314 - [ INFO ] [id: 14] (Logger.logInfo): [OpenSpace] Map 'bpb_E2#E2_Map' already available in cache
2011/02/01 17:45:29.434 - [ INFO ] [id: 14] (Logger.logInfo): [OpenSpace] Map 'bpb_E2#E2_Map' sent to user MarcButton [17]
2011/02/01 17:46:13.640 - [ FINE ] [id: 15] (EventWriter.processEvent): failed writing to channel: java.nio.channels.SocketChannel[connected local=/0.0.0.0:9339 remote=/68.83.11.50:50490]
2011/02/01 17:46:13.650 - [ FINE ] [id: 13] (SmartFoxServer.lostConnection): User [ MarcButton ] removed
2011/02/01 17:46:21.592 - [ FINE ] [id: 16] (EventWriter.dropListOfClients): Disconnecting 1 dead channels.
2011/02/01 17:46:21.592 - [ FINE ] [id: 16] (SmartFoxServer.lostConnection): User [ java.nio.channels.SocketChannel[closed] ] removed

Post by **Lapo** » 02 Feb 2011, 12:05

This error message is of minor importance (it's a FINE level) because it doesn't notify a server error. The message signals that the client on the other end of the socket is not available anymore.
If this problem happens frequently there could be network issues.

Also if you take a look here:

2011/02/01 17:45:29.434 - [ INFO ] [id: 14] (Logger.logInfo): [OpenSpace] Map 'bpb_E2#E2_Map' sent to user MarcButton [17]
2011/02/01 17:46:13.640 - [ FINE ] [id: 15] (EventWriter.processEvent): failed writing to channel: java.nio.channels.SocketChannel[connected local=/0.0.0.0:9339

The two messages are 45 seconds apart. It is possible that the client is idle for too long? How did you configure the <MaxIdleTimer> setting?
Again, on a problematic network a 45 seconds of network traffic pause might cause a loss of connection.

whornak · Post by **whornak** » 02 Feb 2011, 14:36

I could not find a setting called <MaxIdleTimer> in the documentation nor the config file but have the following set in config.xml.

<MaxUserIdleTime>1200</MaxUserIdleTime>
<MaxSocketIdleTime>60</MaxSocketIdleTime>

Post by **Lapo** » 02 Feb 2011, 15:39

Yes, sorry I was going by memory.
Do these error appear systematically for each user? Can you explain the frequency of the problem or maybe you have a way to reproduce it?

Marc · Post by **Marc** » 02 Feb 2011, 22:32

Hey Lapo...an email was sent yesterday to "info" with Wayne on the CC, we would like to try and trouble shoot this deeper as it is occurring on a desktop and laptop at 2 different locations (separate ISP and networks).

Frequency: every time when trying to switch zones, 15-20 seconds and I'm sent back to the login screen as well as the test user on a laptop.

Thanks Again

Post by **Lapo** » 03 Feb 2011, 07:22

Ok, so you can send me a link where I can see this at work?
Also why would switching Zones take 15-20 seconds?
You mean that AFTER having switched Zone you get the problem, roughly 15-20 seconds after you got into the new Zone?

Marc · Post by **Marc** » 03 Feb 2011, 10:33

Wayne will be sending you the link, login information and anything else you would need via email.

Thanks again for the help.

Post by **Lapo** » 03 Feb 2011, 12:34

Sure, let's go ahead by email.

whornak · Post by **whornak** » 03 Feb 2011, 13:25

Email sent @ Thu 2/3/2011 9:25 AM