[ejabberd] ejabberd failure on stress testing

Eric Cestari eric at ohmforce.com
Sun Oct 19 03:10:44 MSD 2008


Seems like you got bitten by the file descriptor limit.

See Jack Moffit's blog post :
http://metajack.im/2008/09/23/file-descriptors-are-yummy-or-common-pitfalls-of-ejabberd/

Eric
Le 19 oct. 08 à 00:58, Arne Claassen a écrit :

> And i hadn't even gotten to the stress part. I'd just connected 1000  
> clients and it was sitting idle with those clients connected when i  
> got this:
>
> =ERROR REPORT==== 2008-10-18 22:50:46 ===
> Mnesia('ejabberd at ip-10-250-11-49'): ** ERROR ** (could not write  
> core file: emfile)
> ** FATAL ** Cannot open log file "/getnotifyd/data/ejabberd/mnesia/ 
> ejabberd at ip-10-250-11-49/DECISION_TAB.TMP": {file_error,
>                                                                                                                 "/getnotifyd 
> /data/ejabberd/mnesia/ejabberd at ip-10-250-11-49/DECISION_TAB.TMP",
>                                                                                                                 emfile 
> }
>
> =INFO REPORT==== 2008-10-18 22:50:46 ===
>    application: mnesia
>    exited: shutdown
>    type: temporary
>
> =ERROR REPORT==== 2008-10-18 22:50:46 ===
> ** Generic server ejabberd_sm terminating
> ** Last message in was {mnesia_system_event,
>                           {mnesia_down,'ejabberd at ip-10-250-11-49'}}
> ** When Server state == {state}
> ** Reason for termination ==
> ** {{function_clause,
>        [{gen,call,
>             [{rex,{error, 
> {node_not_running,'ejabberd at ip-10-250-11-49'}}},
>              '$gen_call',
>              {call,mnesia,remote_dirty_select,
>                  [session,
>                   [{{session,{'_','$1'},'_','_','_','_'},
>                     [{'==',{node,'$1'},'ejabberd at ip-10-250-11-49'}],
>                     ['$_']}]],
>                  <0.36.0>},
>              infinity]},
>         {gen_server,call,3},
>         {rpc,do_call,3},
>         {mnesia,do_dirty_rpc,5},
>         {ejabberd_sm,'-clean_table_from_bad_node/1-fun-1-',1},
>         {mnesia_tm,non_transaction,5},
>         {ejabberd_sm,handle_info,2},
>         {gen_server,handle_msg,5}]},
>    {gen_server,call,
>        [{rex,{error,{node_not_running,'ejabberd at ip-10-250-11-49'}}},
>         {call,mnesia,remote_dirty_select,
>             [session,
>              [{{session,{'_','$1'},'_','_','_','_'},
>                [{'==',{node,'$1'},'ejabberd at ip-10-250-11-49'}],
>                ['$_']}]],
>             <0.36.0>},
>         infinity]}}
>
> =INFO REPORT==== 2008-10-18 22:50:46 ===
> D(<0.6955.0>:ejabberd_receiver:298) : Received XML on stream = " "
>
> =INFO REPORT==== 2008-10-18 22:50:46 ===
> D(<0.8303.0>:ejabberd_receiver:298) : Received XML on stream = " "
>
> =ERROR REPORT==== 2008-10-18 22:50:46 ===
>    application_master: shutdown_error
>    ejabberd_app: {prep_stop,[[]]}
>    error_info: {badarg,[{ets,lookup,[config,hosts]},
>                         {ejabberd_config,get_global_option,1},
>                         {ejabberd_app,stop_modules,0},
>                         {ejabberd_app,prep_stop,1},
>                         {application_master,prep_stop,2},
>                         {application_master,loop_it,4}]}
>
> =INFO REPORT==== 2008-10-18 22:50:46 ===
>    application: ejabberd
>    exited: shutdown
>    type: temporary
>
> Looks like IO.. This is on an EC2 instance with just the regular  
> disk for a small instance. Any suggestions? Gonna try this some  
> more, see if i can replicate the failure, or if this was a freak  
> occurence.
>
> thanks,
> arne
>
>
> _______________________________________________
> ejabberd mailing list
> ejabberd at jabber.ru
> http://lists.jabber.ru/mailman/listinfo/ejabberd



More information about the ejabberd mailing list