shared mem error with 6.0

govtcheeze · #1 10-15-2004, 02:07 PM

Using the new WR 6.0, clean compile of zone, world and sharedmem. This is my world output:

Code:

[Status] Using database 'emu' at localhost
[Status] CURRENT_WORLD_VERSION:EQEMu 0.6.0-DR1
[Status] Loading variables..
[Status] Loading zones..
[Status] Clearing groups..
[Status] Loading items..
[Status] libEMuShareMem.so loaded
[b]Error: EMuShareMem: DLLLoadItems: !ItemsMMF.IsLoaded() (timeout)
[Error] Error: Could not load item data.  But ignoring[/b]
[Status] ./LoginServer.ini read.
[Status] Loading guild ranks..
[Status] Loading ./addon.ini..
[Status] Loading EQ time of day..
[Status] Deleted 0 stale player corpses from database
[Status] Deleted 0 stale player backups from database
Warning: You need to get laid...
TCP listening on: xxx.xxx.xxx.xxx:9000
World server listening on: xxx.xxx.xxx.xxx:9000
Connected to LoginServer: 207.36.180.194:5994

I made sure db.ini is set up correctly with my DB info...ideas?

**Doodman** · #2 10-15-2004, 05:12 PM

I'm assuming on linux.

Stop world and zones.

Check for shared memory segments:

Code:

% ipcs -m

------ Shared Memory Segments --------
key        shmid      owner      perms      bytes      nattch     status      
0x49056d82 4980740    doodman   600        16400909   0

If there is one, and nattach is 0, remove it:

Code:

% ipcrm -m 4980740

Shutting down world/zones doesn't clean up shared memory properly, sometimes you need to remove it yourself.

govtcheeze · #3 10-16-2004, 05:04 AM

Ok found some more info on this problem...

ipcs -m showed nothing listed, so I started world again while showing process lists for sql:

Code:

+----+---------+-----------+----+---------+------+-------+------------------+
| Id | User    | Host      | db | Command | Time | State | Info             |
+----+---------+-----------+----+---------+------+-------+------------------+
| 16 | user | localhost |    | Query   | 1    |       | show processlist |
+----+---------+-----------+----+---------+------+-------+------------------+

+----+----------------------+-----------+----+---------+------+------------------+------------------+
| Id | User                 | Host      | db | Command | Time | State            | Info             |
+----+----------------------+-----------+----+---------+------+------------------+------------------+
| 16 | user              | localhost |    | Query   | 0    |                  | show processlist |
| 17 | unauthenticated user | localhost |    | Connect |      | Reading from net |                  |
+----+----------------------+-----------+----+---------+------+------------------+------------------+

+----+---------+-----------+----------+---------+------+-------+------------------+
| Id | User    | Host      | db       | Command | Time | State | Info             |
+----+---------+-----------+----------+---------+------+-------+------------------+
| 16 | user | localhost |          | Query   | 0    |       | show processlist |
| 17 | user | localhost | eqemu | Sleep   | 0    |       |                  |
+----+---------+-----------+----------+---------+------+-------+------------------+

Then this popped up in shell:

Code:

$ SharedMemory->datasize != iSize, We can rebuild him faster better STRONGER!
Or not.. restart all servers on this machine

And the sql process went away.

Some more info: running Fedora Core 2, the June kernel (upgrading to newest right now), mysql 3.26.58 (also will be upgrading this next). I read some issues on the board with Fedora 2, but they seem to have been taken care of in the src.