[SOLVED] Stuck on retrieving characters list

[darksoke]

I've started to encounter this problem and if happen for many other players on my server , they get stuck at retrieving characters list and i receive this weird errors in Server.log

Received unexpected opcode [CMSG_CANCEL_TRADE 0x11C (284)] Status: STATUS_LOGGEDIN_OR_RECENTLY_LOGGOUT Reason: the player has not logged in yet and not recently logout from [Player: <none> (Guid: 0, Account: 11728)]
WorldSocket::handle_input_header(): client (account: 0, char [GUID: 0, name: <none>]) sent malformed packet (size: 51046, cmd: 1260393683)
Received unexpected opcode [CMSG_CANCEL_TRADE 0x11C (284)] Status: STATUS_LOGGEDIN_OR_RECENTLY_LOGGOUT Reason: the player has not logged in yet and not recently logout from [Player: <none> (Guid: 0, Account: 4324)]
WorldSocket::open: peer().get_remote_addr errno = Transport endpoint is not connected
WorldSocket::open: peer().get_remote_addr errno = Transport endpoint is not connected
WorldSocket::open: peer().get_remote_addr errno = Transport endpoint is not connected

I've never seen them before , anny ideea what could it be ?

 

[Reck]

It sounds like either the core login handler, or the wow client has been altered. Have you made any changes to either the core or the wow client, regarding the login system?

 

[darksoke]

It sounds like either the core login handler, or the wow client has been altered. Have you made any changes to either the core or the wow client, regarding the login system?

no core modification for login handlers and also i'm not the only having this problem

 

[Reck]

If you're using a custom wow.exe file, try using an original one, also without custom patches, and see if you can login then

 

[darksoke]

If you're using a custom wow.exe file, try using an original one, also without custom patches, and see if you can login then

patches does not have annything to do serverside ... that errors are caused by even the CORE itself or DB but i'm lost

 

[Tommy]

Look at: https://github.com/TrinityCore/TrinityCore/issues/558

 

[darksoke]

i don't understand annything from there

 

[Jafferwaffer]

solved.
It was local problem caused by the mess in realmd database...
... feel free to close this issue

From near the bottom of the thread. Check over your db

 

[darksoke]

well it doesn't helped at all but I'll check

 

[Tommy]

well it doesn't helped

So negative. It is something to go by, so yes, it helps since it is basically telling you where to look so it can be potentially fixed. Else, if you aren't finding the help you need here, I would advise you to create a new issue on TrinityCore's issue tracker so they can fix it.

 

[darksoke]

Clean auth and characters DB stil stuck at retrieving chars list even if those errors does not occur but the stuck problems happen after like 1 hour of runing

actually it happen when there are more than 12 players online

RAM usage on host is 15%

Tc won't help me because is an old patched trinitycore , and the the first answer from them is always "we don't support patched or modified version please update to our latest shit bla bla"

 

[darksoke]

well since it look like nobody could help me with annything i think i have found a fix for that but i'm not 100% i'll keep checking the performance all day to see if it got fixed , here is what i've done.

ATTENTION I'VE DONE ON A LINUX BASED SERVER PERFORMANCE CONFIGS ARE NOT THE SAME AS ON WINDOWS.

1st- reinstalled the mysql server because after a flush of locked tables i think i messed up the information_schema
2nd- in authserver.conf and worldserver.conf i set UseProcessors = -15(as specified on TC linux based use -15) and ProcessPriority = 1
3rd- in my.cnf(mysql server config file) i've done the following changes for some reassons the mysql only accepted 10 connections that's why things got messed up after reaching 12 + players on so here are my configs there (My server specifications are the following - hexacore - 16 GB Ram - Ubuntu OS)

key_buffer		= 16M
max_allowed_packet	= 128M  //this was set to 16M by default but as i read on TC 128M can be used if you have more than 2.5GB RAM
max_connections         = 4096  //this one was originally set to 10
table_cache             = 4096  //and this one was set to 1 both by default
query_cache_limit	= 50M   //this if i remember well was set to 1M by default 
query_cache_size        = 512M  //and this one was set to 50M i don't see anny logic on both cache settings values

The reasson why I posted it is maybe it will help others and I'm looking forward for other suggestions for a better performance on the system I use.Plus I was tired of TC shit they keep telling everyone having this issue to update the core to the latest "stable" version but why to do it when this problem appear randomly after 1 year of runing with no problems .. Sorry for bad bad lang I haven't sleep for almost 2 days.

 

[Tommy]

well since it look like nobody could help me with annything

Just because you ask for support doesn't mean you'll get support, especially for an issue that looks to me like it is pretty rare to be receiving. If nobody has a solution, no one is going to reply since it would be quite pointless to. Can't always make the assumption that everyone knows everything anywhere and complain or make snide remarks. Just saying.

It is good you potentially found a fix and we encourage other users to do what you just did after potentially solving an issue: explain in detail how you fixed it (or tried to fix it).

After looking it did seem as if this issue was database related, mainly since the issue I linked above references the database.

TC shit they keep telling everyone having this issue to update the core to the latest "stable" version but why to do it when this problem appear randomly after 1 year of runing with no problems

I agree. However, sometimes they fix issues in future commits, but I would assume only if those said commits were reported (or if some team member had a bright idea to solve some issues).

 

[darksoke]

Just because you ask for support doesn't mean you'll get support, especially for an issue that looks to me like it is pretty rare to be receiving. If nobody has a solution, no one is going to reply since it would be quite pointless to. Can't always make the assumption that everyone knows everything anywhere and complain or make snide remarks. Just saying.

It is good you potentially found a fix and we encourage other users to do what you just did after potentially solving an issue: explain in detail how you fixed it (or tried to fix it).

After looking it did seem as if this issue was database related, mainly since the issue I linked above references the database.



I agree. However, sometimes they fix issues in future commits, but I would assume only if those said commits were reported (or if some team member had a bright idea to solve some issues).

srry didn't mean to be rude i haven't sleep for 2 days trying to fix this issue i only ment to help others and maybe get some advice about a better setup it maybe souds a bit bad but i escused myself for the bad language at the end of the topic didn't mean to sound like that.