Mailing List Archive

(very long) A problem using 0.6.1-pre6 on 2.4.13-smp kernel
I'm trying to install drbd 0.6.1-pre6 on a 2.4.13 kernel, smp, 1Gb ram.

When compiling I see

gcc -c -g -D__KERNEL__ -DMODULE -O2 -Wall -fno-strict-aliasing -mpreferr
ed-stack-boundary=2 -march=i686 -DCPU=686 -Wstrict-prototypes -Wno-trigr
aphs -fomit-frame-pointer -fno-common -pipe -I/usr/src/linux/include -D
__SMP__ -DMODVERSION -DCONFIG_MODVERSION -DEXPORT_SYMTAB -DAPI_VERSION=5
8 -DPRO_VERSION=60 -DREL_VERSION=\"0.6.1-pre6\" drbd_syncer.c
drbd_main.c:953: warning: function declaration isn't a prototype
drbd_main.c:980: warning: function declaration isn't a prototype
/usr/src/linux/include/linux/blk.h:358: warning: `do_nbd_request'
declared `static' but never defined

(the different command line for gcc is a test of mine, I tried the plain
cvs
distribution, too, with identical problem as described below).

and a depmod -a says

depmod: *** Unresolved symbols in
/lib/modules/2.4.13-64Gb-SMP/kernel/drivers/block/drbd.o

but the module load.

I start drbd on two servers (asso1 and asso2) and the sycronization
process finish
without problem. As phisical unit I use /dev/md4 (software raid) on both
servers.
mke2fs /dev/nb0 doesn't report errors.

Using the network disk, however, is almost impossible.
Without using the disk, I find in the syslog these lines:

Oct 26 20:15:00 asso1 kernel: drbd0: sock_sendmsg returned -14
Oct 26 20:15:00 asso1 kernel: drbd0: Connection lost.(pc=1,uc=0)
Oct 26 20:15:00 asso1 kernel: drbd0: asender terminated
Oct 26 20:15:00 asso1 kernel: drbd0: Connection established.
Oct 26 20:15:00 asso1 kernel: drbd0: size=12209280 KB / blksize=4096 B
Oct 26 20:15:00 asso1 kernel: drbd0: Synchronisation started blks=15
int=1
Oct 26 20:15:00 asso1 kernel: drbd0: Synchronisation done.
Oct 26 20:15:34 asso1 kernel: drbd0: sock_sendmsg returned -14
Oct 26 20:15:34 asso1 kernel: drbd0: Connection lost.(pc=1,uc=0)
Oct 26 20:15:34 asso1 kernel: drbd0: asender terminated
Oct 26 20:15:35 asso1 kernel: drbd0: Connection established.
Oct 26 20:15:35 asso1 kernel: drbd0: size=12209280 KB / blksize=4096 B
Oct 26 20:15:35 asso1 kernel: drbd0: Synchronisation started blks=15
int=1
Oct 26 20:15:35 asso1 kernel: drbd0: Synchronisation done.
Oct 26 20:17:05 asso1 kernel: drbd0: ping ack did not arrive
Oct 26 20:17:05 asso1 kernel: drbd0: Connection lost.(pc=0,uc=0)
Oct 26 20:17:05 asso1 kernel: drbd0: asender terminated
Oct 26 20:17:05 asso1 kernel: drbd0: Connection established.
Oct 26 20:17:05 asso1 kernel: drbd0: size=12209280 KB / blksize=4096 B
Oct 26 20:17:05 asso1 kernel: drbd0: Synchronisation started blks=15
int=1
Oct 26 20:17:05 asso1 kernel: drbd0: Synchronisation done.
Oct 26 20:15:31 asso2 kernel: drbd0: rr=0
Oct 26 20:15:31 asso2 kernel: drbd0: asender terminated
Oct 26 20:15:31 asso2 kernel: drbd0: Connection lost.(pc=0,uc=0)
Oct 26 20:15:31 asso2 kernel: drbd0: Connection established.
Oct 26 20:15:31 asso2 kernel: drbd0: size=12209280 KB / blksize=4096 B
Oct 26 20:16:05 asso2 kernel: drbd0: Connection lost.(pc=0,uc=0)
Oct 26 20:16:05 asso2 kernel: drbd0: asender terminated
Oct 26 20:16:06 asso2 kernel: drbd0: Connection established.
Oct 26 20:16:06 asso2 kernel: drbd0: size=12209280 KB / blksize=4096 B
Oct 26 20:17:36 asso2 kernel: drbd0: Connection lost.(pc=0,uc=0)
Oct 26 20:17:36 asso2 kernel: drbd0: asender terminated
Oct 26 20:17:36 asso2 kernel: drbd0: Connection established.
Oct 26 20:17:36 asso2 kernel: drbd0: size=12209280 KB / blksize=4096 B

The disk is only mounted (primary asso1), but not used.

Then I try some write...

asso1:/cluster # time dd if=/dev/zero of=test bs=4k count=1
1+0 records in
1+0 records out

real 0m0.110s
user 0m0.000s
sys 0m0.010s

(no problem)

asso1:/cluster # time dd if=/dev/zero of=test bs=4k count=10
10+0 records in
10+0 records out

real 0m0.458s
user 0m0.000s
sys 0m0.000s

(again, no problem)

asso1:/cluster # time dd if=/dev/zero of=test bs=4k count=100
100+0 records in
100+0 records out

real 0m6.490s
user 0m0.000s
sys 0m0.000s

(ok, but I think it's slow)

asso1:/cluster # time dd if=/dev/zero of=test bs=4k count=1000
1000+0 records in
1000+0 records out

real 0m33.107s
user 0m0.000s
sys 0m0.080s

Here syslog starts to report

Oct 26 20:31:35 asso1 kernel: drbd0: sock_sendmsg returned -14
Oct 26 20:31:35 asso1 kernel: drbd0: Connection lost.(pc=3,uc=0)
Oct 26 20:31:35 asso1 kernel: drbd0: asender terminated
Oct 26 20:31:35 asso1 kernel: drbd0: Connection established.
Oct 26 20:31:35 asso1 kernel: drbd0: size=12209280 KB / blksize=4096 B
Oct 26 20:31:35 asso1 kernel: drbd0: Synchronisation started blks=15
int=1
Oct 26 20:31:35 asso1 kernel: drbd0: Synchronisation done.
Oct 26 20:32:00 asso1 kernel: drbd0: sock_sendmsg returned -14
Oct 26 20:32:00 asso1 kernel: drbd0: Connection lost.(pc=1,uc=0)
Oct 26 20:32:00 asso1 kernel: drbd0: asender terminated
Oct 26 20:32:00 asso1 kernel: drbd0: Connection established.
Oct 26 20:32:00 asso1 kernel: drbd0: size=12209280 KB / blksize=4096 B
Oct 26 20:32:00 asso1 kernel: drbd0: Synchronisation started blks=15
int=1
Oct 26 20:32:00 asso1 kernel: drbd0: Synchronisation done.
Oct 26 20:32:05 asso1 kernel: drbd0: sock_sendmsg returned -14
Oct 26 20:32:05 asso1 kernel: drbd0: Connection lost.(pc=0,uc=0)
Oct 26 20:32:05 asso1 kernel: drbd0: asender terminated
Oct 26 20:32:05 asso1 kernel: drbd0: Connection established.
Oct 26 20:32:05 asso1 kernel: drbd0: size=12209280 KB / blksize=4096 B
Oct 26 20:32:05 asso1 kernel: drbd0: Synchronisation started blks=15
int=1
Oct 26 20:32:05 asso1 kernel: drbd0: Synchronisation done.
Oct 26 20:32:10 asso1 kernel: drbd0: sock_sendmsg returned -14
Oct 26 20:32:10 asso1 kernel: drbd0: Connection lost.(pc=0,uc=0)
Oct 26 20:32:10 asso1 kernel: drbd0: asender terminated
Oct 26 20:32:10 asso1 kernel: drbd0: Connection established.
Oct 26 20:32:10 asso1 kernel: drbd0: size=12209280 KB / blksize=4096 B
Oct 26 20:32:10 asso1 kernel: drbd0: Synchronisation started blks=15
int=1
Oct 26 20:32:10 asso1 kernel: drbd0: Synchronisation done.
Oct 26 20:32:15 asso1 kernel: drbd0: sock_sendmsg returned -14
Oct 26 20:32:15 asso1 kernel: drbd0: Connection lost.(pc=0,uc=0)
Oct 26 20:32:15 asso1 kernel: drbd0: asender terminated
Oct 26 20:32:15 asso1 kernel: drbd0: Connection established.
Oct 26 20:32:15 asso1 kernel: drbd0: size=12209280 KB / blksize=4096 B
Oct 26 20:32:15 asso1 kernel: drbd0: Synchronisation started blks=15
int=1
Oct 26 20:32:15 asso1 kernel: drbd0: Synchronisation done.
Oct 26 20:32:20 asso1 kernel: drbd0: sock_sendmsg returned -14
Oct 26 20:32:20 asso1 kernel: drbd0: Connection lost.(pc=0,uc=0)
Oct 26 20:32:20 asso1 kernel: drbd0: asender terminated
Oct 26 20:32:20 asso1 kernel: drbd0: Connection established.
Oct 26 20:32:20 asso1 kernel: drbd0: size=12209280 KB / blksize=4096 B
Oct 26 20:32:20 asso1 kernel: drbd0: Synchronisation started blks=15
int=1
Oct 26 20:32:20 asso1 kernel: drbd0: Synchronisation done.
Oct 26 20:32:25 asso1 kernel: drbd0: sock_sendmsg returned -14
Oct 26 20:32:25 asso1 kernel: drbd0: Connection lost.(pc=0,uc=0)
Oct 26 20:32:25 asso1 kernel: drbd0: asender terminated
Oct 26 20:32:25 asso1 kernel: drbd0: Connection established.
Oct 26 20:32:25 asso1 kernel: drbd0: size=12209280 KB / blksize=4096 B
Oct 26 20:32:25 asso1 kernel: drbd0: Synchronisation started blks=15
int=1
Oct 26 20:32:25 asso1 kernel: drbd0: Synchronisation done.
Oct 26 20:32:30 asso1 kernel: drbd0: sock_sendmsg returned -14
Oct 26 20:32:30 asso1 kernel: drbd0: Connection lost.(pc=0,uc=0)
Oct 26 20:32:30 asso1 kernel: drbd0: asender terminated
Oct 26 20:32:30 asso1 kernel: drbd0: Connection established.
Oct 26 20:32:30 asso1 kernel: drbd0: size=12209280 KB / blksize=4096 B
Oct 26 20:32:30 asso1 kernel: drbd0: Synchronisation started blks=15
int=1
Oct 26 20:32:30 asso1 kernel: drbd0: Synchronisation done.
Oct 26 20:32:35 asso1 kernel: drbd0: sock_sendmsg returned -14
Oct 26 20:32:35 asso1 kernel: drbd0: Connection lost.(pc=0,uc=0)
Oct 26 20:32:35 asso1 kernel: drbd0: asender terminated
Oct 26 20:32:35 asso1 kernel: drbd0: Connection established.
Oct 26 20:32:35 asso1 kernel: drbd0: size=12209280 KB / blksize=4096 B
Oct 26 20:32:35 asso1 kernel: drbd0: Synchronisation started blks=15
int=1
Oct 26 20:32:35 asso1 kernel: drbd0: Synchronisation done.
Oct 26 20:32:05 asso2 kernel: drbd0: sock_sendmsg returned -104
Oct 26 20:32:05 asso2 kernel: drbd0: Connection lost.(pc=0,uc=0)
Oct 26 20:32:05 asso2 kernel: drbd0: asender terminated
Oct 26 20:32:06 asso2 kernel: drbd0: Connection established.
Oct 26 20:32:06 asso2 kernel: drbd0: size=12209280 KB / blksize=4096 B
Oct 26 20:32:31 asso2 kernel: drbd0: Connection lost.(pc=0,uc=0)
Oct 26 20:32:31 asso2 kernel: drbd0: asender terminated
Oct 26 20:32:31 asso2 kernel: drbd0: Connection established.
Oct 26 20:32:31 asso2 kernel: drbd0: size=12209280 KB / blksize=4096 B
Oct 26 20:32:36 asso2 kernel: drbd0: Connection lost.(pc=0,uc=0)
Oct 26 20:32:36 asso2 kernel: drbd0: asender terminated
Oct 26 20:32:36 asso2 kernel: drbd0: Connection established.
Oct 26 20:32:36 asso2 kernel: drbd0: size=12209280 KB / blksize=4096 B
Oct 26 20:32:41 asso2 kernel: drbd0: Connection lost.(pc=0,uc=0)
Oct 26 20:32:41 asso2 kernel: drbd0: asender terminated
Oct 26 20:32:41 asso2 kernel: drbd0: Connection established.
Oct 26 20:32:41 asso2 kernel: drbd0: size=12209280 KB / blksize=4096 B
Oct 26 20:32:46 asso2 kernel: drbd0: Connection lost.(pc=0,uc=0)
Oct 26 20:32:46 asso2 kernel: drbd0: asender terminated
Oct 26 20:32:46 asso2 kernel: drbd0: Connection established.
Oct 26 20:32:46 asso2 kernel: drbd0: size=12209280 KB / blksize=4096 B
Oct 26 20:32:51 asso2 kernel: drbd0: Connection lost.(pc=0,uc=0)
Oct 26 20:32:51 asso2 kernel: drbd0: asender terminated
Oct 26 20:32:51 asso2 kernel: drbd0: Connection established.
Oct 26 20:32:51 asso2 kernel: drbd0: size=12209280 KB / blksize=4096 B
Oct 26 20:32:56 asso2 kernel: drbd0: Connection lost.(pc=0,uc=0)
Oct 26 20:32:56 asso2 kernel: drbd0: asender terminated
Oct 26 20:32:56 asso2 kernel: drbd0: Connection established.
Oct 26 20:32:56 asso2 kernel: drbd0: size=12209280 KB / blksize=4096 B
Oct 26 20:33:01 asso2 kernel: drbd0: Connection lost.(pc=0,uc=0)
Oct 26 20:33:01 asso2 kernel: drbd0: asender terminated
Oct 26 20:33:01 asso2 kernel: drbd0: Connection established.
Oct 26 20:33:01 asso2 kernel: drbd0: size=12209280 KB / blksize=4096 B
Oct 26 20:33:06 asso2 kernel: drbd0: Connection lost.(pc=0,uc=0)
Oct 26 20:33:06 asso2 kernel: drbd0: asender terminated
Oct 26 20:33:06 asso2 kernel: drbd0: Connection established.
Oct 26 20:33:06 asso2 kernel: drbd0: size=12209280 KB / blksize=4096 B

The problem is not the network, I use a crossover cable on a dedicated
network card to join servers. Also, disks are fast:

asso1:/cluster # time dd if=/dev/zero of=/test bs=4k count=1000
1000+0 records in
1000+0 records out

real 0m0.560s
user 0m0.000s
sys 0m0.180s

Servers are dual pentium 3 processors.

drbd.conf is

resource drbd0 {

protocol=B
fsckcmd=fsck -p -y

disk {
do-panic
disk-size=12209280
}

net {
sync-rate=12000
# here I tried default value, 2500, 5000, 10000 as sync-rate
tl-size=512
# tried default tl-size
timeout=60
connect-int=10
ping-int=10
time=5
}

on asso1 {
device=/dev/nb0
disk=/dev/md4
address=192.168.0.1
port=6004
}

on asso2 {
device=/dev/nb0
disk=/dev/md4
address=192.168.0.2
port=6004
}
}

Could anyone help me to solve this problem?
Excuse me for the long message, but I would to send more info
as possible on the situation.

Thanks
Valerio
Re: (very long) A problem using 0.6.1-pre6 on 2.4.13-smp kernel [ In reply to ]
Your kernel has support for 64GB ram. Hmm, this enables himem support and
device drivers like DRBD have to deal with bounce buffers. -- Sorry this
is simply not done yet.
For now: DRBD only works when himem support is disabled.

PS: It should be rather easy to do, in Rubini's book is a good explanation
about this, I think.

-Philipp

* Valerio Granato <contact@example.com> [011026 20:41]:
> I'm trying to install drbd 0.6.1-pre6 on a 2.4.13 kernel, smp, 1Gb ram.
>
> When compiling I see
>
> gcc -c -g -D__KERNEL__ -DMODULE -O2 -Wall -fno-strict-aliasing -mpreferr
> ed-stack-boundary=2 -march=i686 -DCPU=686 -Wstrict-prototypes -Wno-trigr
> aphs -fomit-frame-pointer -fno-common -pipe -I/usr/src/linux/include -D
> __SMP__ -DMODVERSION -DCONFIG_MODVERSION -DEXPORT_SYMTAB -DAPI_VERSION=5
> 8 -DPRO_VERSION=60 -DREL_VERSION=\"0.6.1-pre6\" drbd_syncer.c
> drbd_main.c:953: warning: function declaration isn't a prototype
> drbd_main.c:980: warning: function declaration isn't a prototype
> /usr/src/linux/include/linux/blk.h:358: warning: `do_nbd_request'
> declared `static' but never defined
>
> (the different command line for gcc is a test of mine, I tried the plain
> cvs
> distribution, too, with identical problem as described below).
>
> and a depmod -a says
>
> depmod: *** Unresolved symbols in
> /lib/modules/2.4.13-64Gb-SMP/kernel/drivers/block/drbd.o
>
> but the module load.
>
> I start drbd on two servers (asso1 and asso2) and the sycronization
> process finish
> without problem. As phisical unit I use /dev/md4 (software raid) on both
> servers.
> mke2fs /dev/nb0 doesn't report errors.
>
> Using the network disk, however, is almost impossible.
> Without using the disk, I find in the syslog these lines:
>
> Oct 26 20:15:00 asso1 kernel: drbd0: sock_sendmsg returned -14
> Oct 26 20:15:00 asso1 kernel: drbd0: Connection lost.(pc=1,uc=0)
> Oct 26 20:15:00 asso1 kernel: drbd0: asender terminated
> Oct 26 20:15:00 asso1 kernel: drbd0: Connection established.
> Oct 26 20:15:00 asso1 kernel: drbd0: size=12209280 KB / blksize=4096 B
> Oct 26 20:15:00 asso1 kernel: drbd0: Synchronisation started blks=15
> int=1
> Oct 26 20:15:00 asso1 kernel: drbd0: Synchronisation done.
> Oct 26 20:15:34 asso1 kernel: drbd0: sock_sendmsg returned -14
> Oct 26 20:15:34 asso1 kernel: drbd0: Connection lost.(pc=1,uc=0)
> Oct 26 20:15:34 asso1 kernel: drbd0: asender terminated
> Oct 26 20:15:35 asso1 kernel: drbd0: Connection established.
> Oct 26 20:15:35 asso1 kernel: drbd0: size=12209280 KB / blksize=4096 B
> Oct 26 20:15:35 asso1 kernel: drbd0: Synchronisation started blks=15
> int=1
> Oct 26 20:15:35 asso1 kernel: drbd0: Synchronisation done.
> Oct 26 20:17:05 asso1 kernel: drbd0: ping ack did not arrive
> Oct 26 20:17:05 asso1 kernel: drbd0: Connection lost.(pc=0,uc=0)
> Oct 26 20:17:05 asso1 kernel: drbd0: asender terminated
> Oct 26 20:17:05 asso1 kernel: drbd0: Connection established.
> Oct 26 20:17:05 asso1 kernel: drbd0: size=12209280 KB / blksize=4096 B
> Oct 26 20:17:05 asso1 kernel: drbd0: Synchronisation started blks=15
> int=1
> Oct 26 20:17:05 asso1 kernel: drbd0: Synchronisation done.
> Oct 26 20:15:31 asso2 kernel: drbd0: rr=0
> Oct 26 20:15:31 asso2 kernel: drbd0: asender terminated
> Oct 26 20:15:31 asso2 kernel: drbd0: Connection lost.(pc=0,uc=0)
> Oct 26 20:15:31 asso2 kernel: drbd0: Connection established.
> Oct 26 20:15:31 asso2 kernel: drbd0: size=12209280 KB / blksize=4096 B
> Oct 26 20:16:05 asso2 kernel: drbd0: Connection lost.(pc=0,uc=0)
> Oct 26 20:16:05 asso2 kernel: drbd0: asender terminated
> Oct 26 20:16:06 asso2 kernel: drbd0: Connection established.
> Oct 26 20:16:06 asso2 kernel: drbd0: size=12209280 KB / blksize=4096 B
> Oct 26 20:17:36 asso2 kernel: drbd0: Connection lost.(pc=0,uc=0)
> Oct 26 20:17:36 asso2 kernel: drbd0: asender terminated
> Oct 26 20:17:36 asso2 kernel: drbd0: Connection established.
> Oct 26 20:17:36 asso2 kernel: drbd0: size=12209280 KB / blksize=4096 B
>
> The disk is only mounted (primary asso1), but not used.
>
> Then I try some write...
>
> asso1:/cluster # time dd if=/dev/zero of=test bs=4k count=1
> 1+0 records in
> 1+0 records out
>
> real 0m0.110s
> user 0m0.000s
> sys 0m0.010s
>
> (no problem)
>
> asso1:/cluster # time dd if=/dev/zero of=test bs=4k count=10
> 10+0 records in
> 10+0 records out
>
> real 0m0.458s
> user 0m0.000s
> sys 0m0.000s
>
> (again, no problem)
>
> asso1:/cluster # time dd if=/dev/zero of=test bs=4k count=100
> 100+0 records in
> 100+0 records out
>
> real 0m6.490s
> user 0m0.000s
> sys 0m0.000s
>
> (ok, but I think it's slow)
>
> asso1:/cluster # time dd if=/dev/zero of=test bs=4k count=1000
> 1000+0 records in
> 1000+0 records out
>
> real 0m33.107s
> user 0m0.000s
> sys 0m0.080s
>
> Here syslog starts to report
>
> Oct 26 20:31:35 asso1 kernel: drbd0: sock_sendmsg returned -14
> Oct 26 20:31:35 asso1 kernel: drbd0: Connection lost.(pc=3,uc=0)
> Oct 26 20:31:35 asso1 kernel: drbd0: asender terminated
> Oct 26 20:31:35 asso1 kernel: drbd0: Connection established.
> Oct 26 20:31:35 asso1 kernel: drbd0: size=12209280 KB / blksize=4096 B
> Oct 26 20:31:35 asso1 kernel: drbd0: Synchronisation started blks=15
> int=1
> Oct 26 20:31:35 asso1 kernel: drbd0: Synchronisation done.
> Oct 26 20:32:00 asso1 kernel: drbd0: sock_sendmsg returned -14
> Oct 26 20:32:00 asso1 kernel: drbd0: Connection lost.(pc=1,uc=0)
> Oct 26 20:32:00 asso1 kernel: drbd0: asender terminated
> Oct 26 20:32:00 asso1 kernel: drbd0: Connection established.
> Oct 26 20:32:00 asso1 kernel: drbd0: size=12209280 KB / blksize=4096 B
> Oct 26 20:32:00 asso1 kernel: drbd0: Synchronisation started blks=15
> int=1
> Oct 26 20:32:00 asso1 kernel: drbd0: Synchronisation done.
> Oct 26 20:32:05 asso1 kernel: drbd0: sock_sendmsg returned -14
> Oct 26 20:32:05 asso1 kernel: drbd0: Connection lost.(pc=0,uc=0)
> Oct 26 20:32:05 asso1 kernel: drbd0: asender terminated
> Oct 26 20:32:05 asso1 kernel: drbd0: Connection established.
> Oct 26 20:32:05 asso1 kernel: drbd0: size=12209280 KB / blksize=4096 B
> Oct 26 20:32:05 asso1 kernel: drbd0: Synchronisation started blks=15
> int=1
> Oct 26 20:32:05 asso1 kernel: drbd0: Synchronisation done.
> Oct 26 20:32:10 asso1 kernel: drbd0: sock_sendmsg returned -14
> Oct 26 20:32:10 asso1 kernel: drbd0: Connection lost.(pc=0,uc=0)
> Oct 26 20:32:10 asso1 kernel: drbd0: asender terminated
> Oct 26 20:32:10 asso1 kernel: drbd0: Connection established.
> Oct 26 20:32:10 asso1 kernel: drbd0: size=12209280 KB / blksize=4096 B
> Oct 26 20:32:10 asso1 kernel: drbd0: Synchronisation started blks=15
> int=1
> Oct 26 20:32:10 asso1 kernel: drbd0: Synchronisation done.
> Oct 26 20:32:15 asso1 kernel: drbd0: sock_sendmsg returned -14
> Oct 26 20:32:15 asso1 kernel: drbd0: Connection lost.(pc=0,uc=0)
> Oct 26 20:32:15 asso1 kernel: drbd0: asender terminated
> Oct 26 20:32:15 asso1 kernel: drbd0: Connection established.
> Oct 26 20:32:15 asso1 kernel: drbd0: size=12209280 KB / blksize=4096 B
> Oct 26 20:32:15 asso1 kernel: drbd0: Synchronisation started blks=15
> int=1
> Oct 26 20:32:15 asso1 kernel: drbd0: Synchronisation done.
> Oct 26 20:32:20 asso1 kernel: drbd0: sock_sendmsg returned -14
> Oct 26 20:32:20 asso1 kernel: drbd0: Connection lost.(pc=0,uc=0)
> Oct 26 20:32:20 asso1 kernel: drbd0: asender terminated
> Oct 26 20:32:20 asso1 kernel: drbd0: Connection established.
> Oct 26 20:32:20 asso1 kernel: drbd0: size=12209280 KB / blksize=4096 B
> Oct 26 20:32:20 asso1 kernel: drbd0: Synchronisation started blks=15
> int=1
> Oct 26 20:32:20 asso1 kernel: drbd0: Synchronisation done.
> Oct 26 20:32:25 asso1 kernel: drbd0: sock_sendmsg returned -14
> Oct 26 20:32:25 asso1 kernel: drbd0: Connection lost.(pc=0,uc=0)
> Oct 26 20:32:25 asso1 kernel: drbd0: asender terminated
> Oct 26 20:32:25 asso1 kernel: drbd0: Connection established.
> Oct 26 20:32:25 asso1 kernel: drbd0: size=12209280 KB / blksize=4096 B
> Oct 26 20:32:25 asso1 kernel: drbd0: Synchronisation started blks=15
> int=1
> Oct 26 20:32:25 asso1 kernel: drbd0: Synchronisation done.
> Oct 26 20:32:30 asso1 kernel: drbd0: sock_sendmsg returned -14
> Oct 26 20:32:30 asso1 kernel: drbd0: Connection lost.(pc=0,uc=0)
> Oct 26 20:32:30 asso1 kernel: drbd0: asender terminated
> Oct 26 20:32:30 asso1 kernel: drbd0: Connection established.
> Oct 26 20:32:30 asso1 kernel: drbd0: size=12209280 KB / blksize=4096 B
> Oct 26 20:32:30 asso1 kernel: drbd0: Synchronisation started blks=15
> int=1
> Oct 26 20:32:30 asso1 kernel: drbd0: Synchronisation done.
> Oct 26 20:32:35 asso1 kernel: drbd0: sock_sendmsg returned -14
> Oct 26 20:32:35 asso1 kernel: drbd0: Connection lost.(pc=0,uc=0)
> Oct 26 20:32:35 asso1 kernel: drbd0: asender terminated
> Oct 26 20:32:35 asso1 kernel: drbd0: Connection established.
> Oct 26 20:32:35 asso1 kernel: drbd0: size=12209280 KB / blksize=4096 B
> Oct 26 20:32:35 asso1 kernel: drbd0: Synchronisation started blks=15
> int=1
> Oct 26 20:32:35 asso1 kernel: drbd0: Synchronisation done.
> Oct 26 20:32:05 asso2 kernel: drbd0: sock_sendmsg returned -104
> Oct 26 20:32:05 asso2 kernel: drbd0: Connection lost.(pc=0,uc=0)
> Oct 26 20:32:05 asso2 kernel: drbd0: asender terminated
> Oct 26 20:32:06 asso2 kernel: drbd0: Connection established.
> Oct 26 20:32:06 asso2 kernel: drbd0: size=12209280 KB / blksize=4096 B
> Oct 26 20:32:31 asso2 kernel: drbd0: Connection lost.(pc=0,uc=0)
> Oct 26 20:32:31 asso2 kernel: drbd0: asender terminated
> Oct 26 20:32:31 asso2 kernel: drbd0: Connection established.
> Oct 26 20:32:31 asso2 kernel: drbd0: size=12209280 KB / blksize=4096 B
> Oct 26 20:32:36 asso2 kernel: drbd0: Connection lost.(pc=0,uc=0)
> Oct 26 20:32:36 asso2 kernel: drbd0: asender terminated
> Oct 26 20:32:36 asso2 kernel: drbd0: Connection established.
> Oct 26 20:32:36 asso2 kernel: drbd0: size=12209280 KB / blksize=4096 B
> Oct 26 20:32:41 asso2 kernel: drbd0: Connection lost.(pc=0,uc=0)
> Oct 26 20:32:41 asso2 kernel: drbd0: asender terminated
> Oct 26 20:32:41 asso2 kernel: drbd0: Connection established.
> Oct 26 20:32:41 asso2 kernel: drbd0: size=12209280 KB / blksize=4096 B
> Oct 26 20:32:46 asso2 kernel: drbd0: Connection lost.(pc=0,uc=0)
> Oct 26 20:32:46 asso2 kernel: drbd0: asender terminated
> Oct 26 20:32:46 asso2 kernel: drbd0: Connection established.
> Oct 26 20:32:46 asso2 kernel: drbd0: size=12209280 KB / blksize=4096 B
> Oct 26 20:32:51 asso2 kernel: drbd0: Connection lost.(pc=0,uc=0)
> Oct 26 20:32:51 asso2 kernel: drbd0: asender terminated
> Oct 26 20:32:51 asso2 kernel: drbd0: Connection established.
> Oct 26 20:32:51 asso2 kernel: drbd0: size=12209280 KB / blksize=4096 B
> Oct 26 20:32:56 asso2 kernel: drbd0: Connection lost.(pc=0,uc=0)
> Oct 26 20:32:56 asso2 kernel: drbd0: asender terminated
> Oct 26 20:32:56 asso2 kernel: drbd0: Connection established.
> Oct 26 20:32:56 asso2 kernel: drbd0: size=12209280 KB / blksize=4096 B
> Oct 26 20:33:01 asso2 kernel: drbd0: Connection lost.(pc=0,uc=0)
> Oct 26 20:33:01 asso2 kernel: drbd0: asender terminated
> Oct 26 20:33:01 asso2 kernel: drbd0: Connection established.
> Oct 26 20:33:01 asso2 kernel: drbd0: size=12209280 KB / blksize=4096 B
> Oct 26 20:33:06 asso2 kernel: drbd0: Connection lost.(pc=0,uc=0)
> Oct 26 20:33:06 asso2 kernel: drbd0: asender terminated
> Oct 26 20:33:06 asso2 kernel: drbd0: Connection established.
> Oct 26 20:33:06 asso2 kernel: drbd0: size=12209280 KB / blksize=4096 B
>
> The problem is not the network, I use a crossover cable on a dedicated
> network card to join servers. Also, disks are fast:
>
> asso1:/cluster # time dd if=/dev/zero of=/test bs=4k count=1000
> 1000+0 records in
> 1000+0 records out
>
> real 0m0.560s
> user 0m0.000s
> sys 0m0.180s
>
> Servers are dual pentium 3 processors.
>
> drbd.conf is
>
> resource drbd0 {
>
> protocol=B
> fsckcmd=fsck -p -y
>
> disk {
> do-panic
> disk-size=12209280
> }
>
> net {
> sync-rate=12000
> # here I tried default value, 2500, 5000, 10000 as sync-rate
> tl-size=512
> # tried default tl-size
> timeout=60
> connect-int=10
> ping-int=10
> time=5
> }
>
> on asso1 {
> device=/dev/nb0
> disk=/dev/md4
> address=192.168.0.1
> port=6004
> }
>
> on asso2 {
> device=/dev/nb0
> disk=/dev/md4
> address=192.168.0.2
> port=6004
> }
> }
>
> Could anyone help me to solve this problem?
> Excuse me for the long message, but I would to send more info
> as possible on the situation.
>
> Thanks
> Valerio
>
>
> _______________________________________________
> DRBD-devel mailing list
> DRBD-devel@example.com
> https://lists.sourceforge.net/lists/listinfo/drbd-devel
Re: (very long) A problem using 0.6.1-pre6 on 2.4.13-smp kernel [ In reply to ]
----- Original Message -----
From: "Philipp Reisner" <philipp.reisner@example.com>
To: "Valerio Granato" <contact@example.com>
Cc: <drbd-devel@example.com>
Sent: Saturday, October 27, 2001 8:20 AM
Subject: Re: [DRBD-dev] (very long) A problem using 0.6.1-pre6 on
2.4.13-smp kernel


> Your kernel has support for 64GB ram. Hmm, this enables himem support
and
> device drivers like DRBD have to deal with bounce buffers. -- Sorry
this
> is simply not done yet.

Ok, now I don't see disconnections in the syslog :-)

Only two more questions:

- It's ok a 5MB transfer rate on a dedicated 100Mbit eth?
- Why I find this in the syslog?

Oct 28 09:31:38 asso1 kernel: drbd0: transferlog too small!!
Oct 28 09:31:38 asso1 last message repeated 4 times
Oct 28 09:31:38 asso1 kernel: drbd0: tl messed up!
Oct 28 09:31:38 asso1 kernel: drbd0: Epoch set size wrong!!found=217
reported=1240
Oct 28 09:31:38 asso1 kernel: drbd0: transferlog too small!!
Oct 28 09:31:39 asso1 last message repeated 9 times
Oct 28 09:31:39 asso1 kernel: drbd0: Epoch set size wrong!!found=193
reported=2752
Oct 28 09:31:39 asso1 kernel: drbd0: transferlog too small!!
Oct 28 09:31:39 asso1 kernel: drbd0: Epoch set size wrong!!found=33
reported=288
Oct 28 09:31:40 asso1 kernel: drbd0: transferlog too small!!
Oct 28 09:31:40 asso1 last message repeated 8 times

And so on...

Thanks,
Valerio
Re: (very long) A problem using 0.6.1-pre6 on 2.4.13-smp kernel [ In reply to ]
Increase the tl-size value in drbd.conf, until theese messages diappear.
Start and 6000 and increase it in units of 2000

-Philipp

* Valerio Granato <contact@example.com> [011028 09:40]:
> ----- Original Message -----
> From: "Philipp Reisner" <philipp.reisner@example.com>
> To: "Valerio Granato" <contact@example.com>
> Cc: <drbd-devel@example.com>
> Sent: Saturday, October 27, 2001 8:20 AM
> Subject: Re: [DRBD-dev] (very long) A problem using 0.6.1-pre6 on
> 2.4.13-smp kernel
>
>
> > Your kernel has support for 64GB ram. Hmm, this enables himem support
> and
> > device drivers like DRBD have to deal with bounce buffers. -- Sorry
> this
> > is simply not done yet.
>
> Ok, now I don't see disconnections in the syslog :-)
>
> Only two more questions:
>
> - It's ok a 5MB transfer rate on a dedicated 100Mbit eth?
> - Why I find this in the syslog?
>
> Oct 28 09:31:38 asso1 kernel: drbd0: transferlog too small!!
> Oct 28 09:31:38 asso1 last message repeated 4 times
> Oct 28 09:31:38 asso1 kernel: drbd0: tl messed up!
> Oct 28 09:31:38 asso1 kernel: drbd0: Epoch set size wrong!!found=217
> reported=1240
> Oct 28 09:31:38 asso1 kernel: drbd0: transferlog too small!!
> Oct 28 09:31:39 asso1 last message repeated 9 times
> Oct 28 09:31:39 asso1 kernel: drbd0: Epoch set size wrong!!found=193
> reported=2752
> Oct 28 09:31:39 asso1 kernel: drbd0: transferlog too small!!
> Oct 28 09:31:39 asso1 kernel: drbd0: Epoch set size wrong!!found=33
> reported=288
> Oct 28 09:31:40 asso1 kernel: drbd0: transferlog too small!!
> Oct 28 09:31:40 asso1 last message repeated 8 times
>
> And so on...
>
> Thanks,
> Valerio
>
>
>
> _______________________________________________
> DRBD-devel mailing list
> DRBD-devel@example.com
> https://lists.sourceforge.net/lists/listinfo/drbd-devel