[op5-users] Merlin crashed on me?

[Frater, Greg J]

>> Crash #1
>> daemon.log
>> [1246458609] 7: select() returned 1 (errno = 0: Success) [1246458609]

>> 6: inbound data available on ipc socket [1246458609] 7: Successfully 
>> read 1 NEBCALLBACK_PROGRAM_STATUS_DATA event (352 bytes; 288 bytes 
>> body) from socket 7 [1246458609] 7: sel_val: 7; ipc_listen_sock: 5; 
>> ipc_sock: 7; net_sock: 6 [1246458611] 7: select() returned 1 (errno =

>> 0: Success) [1246458611] 6: inbound data available on ipc socket 
>> [1246458611] 7: Successfully read 1 NEBCALLBACK_HOST_CHECK_DATA event
>> (546 bytes; 482 bytes body) from socket 7 [1246458611] 7: sel_val: 7;

>> ipc_listen_sock: 5; ipc_sock: 7; net_sock: 6 [1246458611] 7: select()

>> returned 1 (errno = 0: Success) [1246458611] 6: inbound data
available 
>> on ipc socket [1246458611] 7: Successfully read 1 
>> NEBCALLBACK_HOST_CHECK_DATA event
>> (486 bytes; 422 bytes body) from socket 7
>> 
>> 
>> Crash #2
>> [1246462221] 7: select() returned 1 (errno = 0: Success) [1246462221]

>> 6: inbound data available on ipc socket [1246462221] 7: Successfully 
>> read 1 NEBCALLBACK_HOST_CHECK_DATA event
>> (546 bytes; 482 bytes body) from socket 7 [1246462221] 7: sel_val: 7;

>> ipc_listen_sock: 5; ipc_sock: 7; net_sock: 6 [1246462221] 7: select()

>> returned 1 (errno = 0: Success) [1246462221] 6: inbound data
available 
>> on ipc socket [1246462221] 7: Successfully read 1 
>> NEBCALLBACK_SERVICE_CHECK_DATA event
>> (575 bytes; 511 bytes body) from socket 7 [1246462221] 7: sel_val: 7;

>> ipc_listen_sock: 5; ipc_sock: 7; net_sock: 6 [1246462221] 7: select()

>> returned 1 (errno = 0: Success) [1246462221] 6: inbound data
available 
>> on ipc socket [1246462221] 7: Successfully read 1 
>> NEBCALLBACK_HOST_CHECK_DATA event
>> (486 bytes; 422 bytes body) from socket 7
>> 
>
>Ooh, I'd quite like to know what that host check looks like. It seems
as if it crashes on the same host-check result both times (judging by
the size only, which is quite a poor heuristic, but still).
>
>I'll re-enable the debugging machinery that dumps inbound messages to a
binary logfile. When that's done, I'll need you to run Merlin until it
crashes again so I get the sequence of events leading up to the actual
crash in the format Merlin sees them. If I replay the same event-chain
on our 64-bit machine, I *should* get the same crash you're getting. If
that's the case, finding and fixing this bug should be fairly trivial.
>

Let me know when you've got that done, I'll run it again.

-greg