ggatec hanging on 7.0-RC1/amd64

[Hugo Silva]

Hi,

I'm experiencing some problems while attempting to export a ZVOL with 
ggated on a gigabit network.


All servers are running FreeBSD 7.0-RC1 on amd64, and using bce NICs.



client# ggatec create -t 6 -u 0 172.16.100.200 /dev/zvol/storage/vol0

client# ls -l /dev/ggate*
crw-r-----  1 root  operator    0, 115 Jan  4 06:23 /dev/ggate0

client# newfs /dev/ggate0
load: 0.00  cmd: newfs 51187 [g_waitidle] 0.00u 0.00s 0% 792k
^C^C^C^C^C^C^C^C^C^C^C^C^C^C


The same happens with a file exported on a UFS filesystem (so it's not a 
ZFS problem):


client# ggatec create -u100 172.16.100.200 /usr/lol

client# dir /dev/ggate100
crw-r-----  1 root  operator  -   0, 115 Jan  4 06:23 /dev/ggate100

client# newfs /dev/ggate100
load: 0.00  cmd: newfs 52306 [g_waitidle] 0.00u 0.00s 0% 792k
load: 0.00  cmd: newfs 52306 [g_waitidle] 0.00u 0.00s 0% 792k
load: 0.00  cmd: newfs 52306 [g_waitidle] 0.00u 0.00s 0% 792k
^C^C^C^C^C^C^C^C^C^C^C^C^C^C^C^C^C^C^C^C^C^C^C^C^C^C^C^C^C^Z^Z^Z^Z^Z^Z^Z^Z^Z^Z^Z^Z^Z^Z^Z^Z^Z^Z^Z^Z


After several killall -9 ggatec and ggatec -fuX (-uX will stop working 
after awhile with "device busy", even though no ggatec process is using 
it anymore and the newfs process is also killed), the following happens:

# killall -9 ggatec
load: 0.00  cmd: killall 52400 [sysctl lock] 0.00u 0.00s 0% 244k
load: 0.00  cmd: killall 52400 [sysctl lock] 0.00u 0.00s 0% 244k
load: 0.00  cmd: killall 52400 [sysctl lock] 0.00u 0.00s 0% 244k

At this point, the machine is dead. It still replies to pings, but 
processes are definitely unkillable now and even trying to login via SSH 
doesn't work:

$ ssh client
load: 0.05  cmd: ssh 2127 [sbwait] 0.00u 0.00s 0% 3180k
[hangs forever]


/etc/gg.exports on the server:
client.ip/32 RW /dev/zvol/storage/vol0


Am I missing something terribly obvious ? I have done this many times 
before and never encountered this problem..

Regards,

Hugo