i have a very stable connection with plenty of bandwidth spare thus plenty of CPU spare so my computer is not generaly over worked but i cant explain why users keep dropping out of my hub all the time?
quite a few of my users have reported this to me since ive changed to 0.3.5.2. ive also been told from several other hub owners that have changed to 0.3.5.2 that they also have the same problem and that they have rolled back to an earlier version to avoid this problem.
i tested this myself by connecting to my hub from another location several miles from my home and i was disconnected 3 times in 2 hours for no reason. i posted this here incase anyone else was having the same problems as me.
thanks
could you provide some more info please ?
What is your total bandwith download/upload ?
What is your O/S, memory and cpu + usage by PtokaX ?
What is that "previous" version that worked better ? (exact version name plz)
How many users does the hub have by average ?
Are there any other hubs running ?
This would help in determining if there's a problem
I have 7 hubs running with the latest ptokax release.
I got disconnected from no one of them and also users don't complain about disconnection at the moment.
You can run wireshark (http://www.wireshark.org/) to get what's happening.
Maybe PtokaX UDP-Debug Receiver (http://www.ptokax.org/files/PtokaX-UDP-Debug-Receiver.7z) show reason why they are disconnected... most of disconnects is because connection reset (WSAECONNRESET) ::)
I have also had the same probs uses disconnect ... but i also had this on 0.3.5.1 . (gone back to 0.3.4.0 )
10MB connection Unsure on download/upload but told unlimited downloads if that helps
Win XP ,P4 ,1024 ram ,5% CPU usage by ptokax .
Previous stable version for me is 0.3.4.0
Average users count is 70 - 90
Am now running PtokaX UDP-Debug Receiver to see if i can narrow down prob/solution
i also gone back to 0.3.5.0 lua5.1 and notice more users in with less disconnects and also notice that the latest version hangs a lot when entering compared to the 0.3.5.0 so that suggest that there is a problem with the latest release
Quoteso that suggest that there is a problem with the latest release
There has been some changes towards LUA libraries as well. So a problem could also be caused by a script. Best way is to test without any running scripts. I myself, do not have any problems with the 0.3.5.2
But like PPK said, that Debugger will show information on why disconnects occur, so please use it and see.
Hi
What does recv() error WSAECONNRESET. User is being closed. mean if there in several hubs and only getting disconnected in this one.
WSAECONNRESET
10054
Quote
Connection reset by peer.
An existing connection was forcibly closed by the remote host. This normally results if the peer application on the remote host is suddenly stopped, the host is rebooted, the host or remote network interface is disabled, or the remote host uses a hard close (see setsockopt for more information on the SO_LINGER option on the remote socket). This error may also result if a connection was broken due to keep-alive activity detecting a failure while one or more operations are in progress. Operations that were in progress fail with WSAENETRESET. Subsequent operations fail with WSAECONNRESET.
For a complete overview of Windows Socket Error Codes, click
http://msdn2.microsoft.com/en-us/library/ms740668.aspx (http://msdn2.microsoft.com/en-us/library/ms740668.aspx)
Basically there are alot of causes for a WSAECONNRESET to occur.
A communication interruption between server and host for example will cause a WSAECONNRESET.
This interruption can be caused by insufficient bandwith, ISP routing issue, 1 of the communicating clients, the operating system, routers, or simply a lost package.
I'm sure there's more reasons to it, and I'm no WinSock expert.
I do know that it occurs, and I do know it occurs regardless of hubsofts
Quote from: CrazyGuy on 16 November, 2006, 18:19:37
WSAECONNRESET
10054
For a complete overview of Windows Socket Error Codes, click
http://msdn2.microsoft.com/en-us/library/ms740668.aspx (http://msdn2.microsoft.com/en-us/library/ms740668.aspx)
Basically there are alot of causes for a WSAECONNRESET to occur.
A communication interruption between server and host for example will cause a WSAECONNRESET.
This interruption can be caused by insufficient bandwith, ISP routing issue, 1 of the communicating clients, the operating system, routers, or simply a lost package.
I'm sure there's more reasons to it, and I'm no WinSock expert.
I do know that it occurs, and I do know it occurs regardless of hubsofts
that doesnt explain why it happens more with the last release of PtokaX [0.3.5.2] so that link would be void in this perticular matter although every release is definatly getting better.
my starting this thread was not to complain but just to bring attention to this matter and to see if i was the only one that was having the same issue with this particular version. anyway i will try that app as PPK suggested and i will report back with my findings.
thanks
ruler
Sorry for the confusion ruler. My answer was to xvk 's question. Maybe I should have qouted him ;)
But yea, please posts some results. It would help to find out if/why this version has more disconnects than others
Quote from: PPK on 15 November, 2006, 16:01:50
Maybe PtokaX UDP-Debug Receiver (http://www.ptokax.org/files/PtokaX-UDP-Debug-Receiver.7z) show reason why they are disconnected... most of disconnects is because connection reset (WSAECONNRESET) ::)
i notice there were no instructions on how to use this app, does it this app need to be run from any spacific folder or can it be run from anywhere? also are the ports on which it listens important?
thanks
Quote from: ruler on 17 November, 2006, 11:43:12
i notice there were no instructions on how to use this app, does it this app need to be run from any spacific folder or can it be run from anywhere? also are the ports on which it listens important?
thanks
The Debugger itself can be run from anywhere. The port is in so far important as it should not be in use yet. The default port is 422.
Select a port and click on Listen.
Then go to the mainchat of your hub and type !debug <port in debugger> on ( by default !debug 422 on )
You will then receive a message back from the hubsoft:
Quote*** You are now subscribed for Debug messaging on UDP port 422. To unsubscribe type !debug off in mainchat.
Go back to the Debugger and you can see that an extra tab named as your hubname has opened.
In there, you will receive the information.
yup thats working now and thanks
ive noticed only after a few seconds of enabling that there is already loads of peeps being disconnected ???
11:51:02 - [HUB] Subscribed.
11:51:02 - [HUB] Users online: 101
11:51:02 - [SYS] New Debug subscriber: [RAW]Owner.ruler (84.9.150.83)
11:51:56 - [SYS] Login timeout (3) 120 seconds for CAS_1980 (82.153.29.192) - user disconnected.
11:52:07 - [SYS] Login timeout (3) 120 seconds for WildboY (213.100.134.147) - user disconnected.
11:52:19 - [SYS] Ghost crispy (84.92.185.142) closed.
11:52:22 - [SYS] Login timeout (3) 120 seconds for Eric_Red (82.67.176.231) - user disconnected.
11:52:44 - [SYS] Nick taken [xy{crocket} (82.38.60.79)] xy{crocket} (82.38.60.79) - user closed.
strange
15:47:57 - [ERR] munter_82 (81.230.196.212): recv() error WSAECONNRESET. User is being closed.
15:48:31 - [SYS] Login timeout (3) 120 seconds for WildboY (213.100.134.147) - user disconnected.
15:48:41 - [SYS] Login timeout (3) 120 seconds for 47??tud!o? (87.74.5.136) - user disconnected.
15:48:50 - [SYS] Login timeout (3) 120 seconds for james (85.217.34.78) - user disconnected.
15:49:02 - [SYS] Login timeout (3) 120 seconds for Eric_Red (82.67.176.231) - user disconnected.
15:49:06 - [SYS] Login timeout 20 seconds for 89.0.6.238 - user disconnected.
15:49:30 - [SYS] Login timeout (3) 120 seconds for Pawian77777 (83.22.112.129) - user disconnected.
15:49:41 - [SYS] Login timeout (3) 120 seconds for mol0tov (217.23.249.83) - user disconnected.
15:50:02 - [SYS] Login timeout (3) 120 seconds for CAS_1980 (82.153.29.192) - user disconnected.
15:50:20 - [SYS] Nick taken [xy{crocket} (82.38.60.79)] xy{crocket} (82.38.60.79) - user closed.
15:51:13 - [SYS] Login timeout (3) 120 seconds for Eric_Red (82.67.176.231) - user disconnected.
15:51:17 - [SYS] Login timeout (3) 120 seconds for WildboY (213.100.134.147) - user disconnected.
15:51:31 - [SYS] Login timeout (3) 120 seconds for 47??tud!o? (87.74.5.136) - user disconnected.
15:51:34 - [SYS] Login timeout (3) 120 seconds for Pawian77777 (83.22.112.129) - user disconnected.
15:51:41 - [REG] Hub is registered on dreamland.gotdns.org hublist (Users: 89, Share: 30824085908498)
15:51:41 - [REG] RegSock connect error WSAECONNREFUSED (10061). (reg.hublist.org)
15:51:47 - [REG] Hub is registered on hubs.hublist.co.uk hublist (Users: 89, Share: 30824085908498)
15:51:48 - [REG] RegSock connect error WSAECONNREFUSED (10061). (bwhubreg.no-ip.info)
15:51:50 - [REG] Hub is registered on reg.ancient.pl hublist (Users: 89, Share: 30824085908498)
15:51:50 - [SYS] Login timeout (3) 120 seconds for mol0tov (217.23.249.83) - user disconnected.
15:51:58 - [REG] RegSock connect error WSAETIMEDOUT (10060). (zengeday.sytes.net)
15:52:07 - [REG] RegSock connect error WSAETIMEDOUT (10060). (hublist.uknnet.com)
15:52:15 - [SYS] Login timeout (3) 120 seconds for CAS_1980 (82.153.29.192) - user disconnected.
15:53:07 - [SYS] Nick taken [xy{crocket} (82.38.60.79)] xy{crocket} (82.38.60.79) - user closed.
15:53:23 - [SYS] Login timeout (3) 120 seconds for Eric_Red (82.67.176.231) - user disconnected.
15:53:27 - [LOG] Cannot allocate memory in UserPutInSendBuf!
15:53:27 - [BUF] kalle507 (213.66.19.17): Re-allocation of sendbuf failed in 'sendbuflen < sbdatalen+iTxtLen' condition.
15:53:27 - [LOG] Serving stopped (UL: 3491039 [193232] , DL: 188111)
15:53:27 - [ERR] accept() has returned an INVALID_SOCKET.
15:53:27 - [ERR] accept() has returned an INVALID_SOCKET.
15:53:27 - [ERR] accept() has returned an INVALID_SOCKET.
15:53:27 - [ERR] accept() has returned an INVALID_SOCKET.
then hub crashed disconnecting everyone. for some reason when running that app that PPK suggested it seems to crash the hub all the time. earlier my hub was running at 100% cpu and wouldnt respond at all, i had to restart my computer to get it all working again ???
Quoteearlier my hub was running at 100% cpu and wouldnt respond at all
You wouldn't happen to have Norton Internet Security running would you ? That caused that problem for me on hub crash
Quotefor some reason when running that app that PPK suggested it seems to crash the hub all the time
I had that problem too. After a first crash, it keeps crashing on hub-startup with the Debugger active..But since I was the only 1 with that problem and I got specific memory allocation errors, we came to the conclusion it was my computer screwing up :P So now I'm not too sure about that either.
The rest of the information seems valuable to me, so let's see what PPK can do with it :) Thanks
i have two main rules which i ALWAYS abide by, i NEVER install anything norton related and NEVER install anything yahoo related. hope that clears that matter up ;D
im using a normal modem with only XP firewall and McAfee AV nothing more. here are my hub stats ...
<[=R=D=]Bot>
------------------------------------------------------------
Current stats:
------------------------------------------------------------
Version: PtokaX DC Hub 0.3.5.2 built on Nov 1 2006 11:01:55
Uptime: 0 days, 1 hours, 27 minutes
Users (Max/Actual Peak (Max Peak)/Logged): 400 / 101 (153) / 100
Users shared size: 32840752886751 Bytes / 29.87 TB
Chat messages: 17 x
Unknown commands: 0 x
PM commands: 0 x
Key commands: 482 x
Supports commands: 512 x
MyINFO commands: 685 x
ValidateNick commands: 448 x
GetINFO commands: 0 x
Password commands: 87 x
Version commands: 223 x
GetNickList commands: 249 x
Search commands: 1274 x (0 x)
SR commands: 814 x
CTM commands: 6528 x (0 x)
RevCTM commands: 3229 x
BotINFO commands: 0 x
------------------------------------------------------------
CPU usage (60 sec avg): 0.15%
CPU time: 0:00:06
Mem usage (Peak): 6.59 MB (9.84 MB)
VM size (Peak): 6.58 MB (6.89 MB)
------------------------------------------------------------
SendRests (Peak): 0 (0)
RecvRests (Peak): 0 (2)
Compression saved: 472.25 kB (43)
Data sent: 15.53 MB
Data received: 771.11 kB
Tx (60 sec avg): 6.66 kB/s (4.50 kB/s)
Rx (60 sec avg): 114 B/s (148 B/s)
my PC is not over loaded in anyway by either CPU, memory or internet bandwidh and the problem only started 'after' i upgraded to ver [0.3.5.2] the logs are already massive on [registered] users that are being disconnected all the time. its more than i knew about
Quote from: ruler on 17 November, 2006, 18:28:11
the logs are already massive on [registered] users that are being disconnected all the time
As i see in you log, they are disconnected in state 3... internal name for state number 3 is STATE_VERSION_OR_MYPASS... maybe they don't enter password when hub asked it ???
they get in eventually but then later they disconnect again. i had to uninstall that app cos it kept crashing the hub all the time and with it running the user count was seriously lower than normal (25% lower)
im wondering if it might be crappy clients/hack versions that it has a problem with? if that is the case then im happy with it disconnecting them ;D