Post: #1

Most of this is written for the developers to look through as I think there may be an issue with the network protocol on the Linux nightly build and, I also have a couple suggestions to go along with that...

1. Shard1 built and testing on VMware 4.1 Cluster on a Windows 2008 R2 64bit OS. Fully patched, VMXNET Ethernet device. While fairly stable, there are multiple network "debugs" that show on the console occasionally but never "shard ending". The server console will stop responding occasionally and hang, causing a full crash.
2. Shard2 built and testing on CentOS 6.4 64bit, VMXNET driver, same VMware cluster. This is the WORST one to try to build as 32 bit dependancies are not fun or easy to do on a 64bit OS. Hard crashes if more than one account logged in, always networking related in the console. BUT, immediately after the "network crash" I can hit #<enter> and the world will save. No traffic is flowing over eth0 however.
3. Shard3 built and testing on CentOS 6.4 32bit, VMXNET driver, same VMware cluster. Almost identical to number 2 above.
4. Hardware server test is underway and will update that when complete.

Now, what I have noticed is, from what I can tell, the VMXNET driver is not liked by SPHERESVR(.EXE). I have not tried an Intel Pro 100/1000 in a VMware machine yet so I am not sure if it is limited to that, but I would doubt it as that is all controlled by the OS, not the SPHERESVR.

What I can ask though is for the devs to PLEASE, PLEASE, PLEASE, update to the latest MySQL library on the Linux builds (16 is the latest that is stable) and allow the server app to pull from either /usr/lib/ or /usr/lib64/

My two cents. Been away from this community to long, and glad to be back!

MPGRAM