[ENBD] (no subject)
Travis, Mark
enbd@lists.community.tummy.com
Thu, 19 Jun 2003 17:12:26 -0700
Regarding Oracle/Dell/RHAS and ENBD, I have a couple of thoughts.
First, I don't work with Oracle cluster. But I do know that vendors of
stuff like that are extremely picky when it comes to the various =
elements of
the architecture. I would be very surprised if Oracle would still want =
to
honor your support agreement if you incorporated ENBD within the =
cluster.
It's nothing against ENBD, which is great. It's more a matter that =
it's
probably an unknown to Oracle -- and they're probably not going to want =
to
support any unknowns--especially in regards to the shared storage which =
is
absolutely crucial to availability and data integrity. So while I'm
speculating, it's probably smart to verify that you'll end up with =
vendor
support if you get to the point of deploying ENBD into production.
Again, I don't know what Oracle would support. And it's probably smart =
of
you not to buy a full-blown SAN which is I'm sure what Dell tried to =
sell
you. You can probably get away with a pair of JBOD SCSI boxes for a =
few
$thousand apiece depending on your requirements and get Oracle support =
(as
long as they validate the JBOD vendor).
I'm not trying to scare you off of ENBD -- it's just that even if it =
works
for you that doesn't guarantee that your vendors will support the
environment if something does go wrong.
-----Original Message-----
From: Javier Llorente [mailto:nibble@jet.es]=20
Sent: Thursday, June 19, 2003 4:17 PM
To: enbd@lists.community.tummy.com
Subject: RE: [ENBD] (no subject)
Here we are... First at all, I=B4ll explain what I want to do; and =
what I=B4ve
done nowadays.
All the tests I=B4ve done with nbd (not enbd) have been only using one
node
working as server and as two clients, mounting the same device in to
differents /dev/ndX, with the device mounted (/dev/hdd4), and sharing =
it
through samba. All rigth, just testing, maybe to much for enbd (I =
don=B4t
know), but ... despite of the deadlocks that halt the system (It =
isn=B4t a
good idea to mount the device in the node that owns it), ... I was able =
to
create a half of the database (400 megas).
I need to mount the same device (read/write) in two diferents nodes
(If
it=B4s possible, the cluster software and the cluster filesystem should =
do the
rest). I=B4ve download the nbd-2.4.30-current version of enbd and untar =
it.
total 125
-rw-r--r-- 1 root root 562 nov 13 2002 =
compile-linux-2.5
-rw-r--r-- 1 root root 64360 mar 27 18:08 DIARY
drwxr-xr-x 2 root root 1024 mar 27 20:11 doc
lrwxrwxrwx 1 root root 12 jun 20 00:33 FILES ->
FILES-2.2.10
-rw-r--r-- 1 root root 317 sep 17 2000 FILES-2.0.36
-rw-r--r-- 1 root root 278 oct 14 2002 FILES-2.2.10
-rw-r--r-- 1 root root 1309 abr 29 2001 FILES-2.4.0
-rw-r--r-- 1 root root 7534 nov 27 2002 INSTALL
-rw-r--r-- 1 root root 17845 sep 17 2000 LICENCE
lrwxrwxrwx 1 root root 12 jun 20 00:33 linux ->
linux-2.2.10
drwxr-xr-x 6 root root 1024 ene 7 2000 linux-2.0.36
drwxr-xr-x 4 root root 1024 ene 7 2000 linux-2.2.10
drwxr-xr-x 4 root root 1024 abr 4 2002 linux-2.4.18
drwxr-xr-x 4 root root 1024 ago 22 2002 linux-2.5.12
drwxr-xr-x 4 root root 1024 dic 11 2001 linux-2.5.31
drwxr-xr-x 4 root root 1024 dic 11 2001 linux-2.5.44
drwxr-xr-x 4 root root 1024 dic 11 2001 linux-2.5.47
drwxr-xr-x 4 root root 1024 dic 11 2001 linux-2.5.7
-rw-r--r-- 1 root root 0 jun 20 00:35 ls.txt
-rwxr-xr-x 1 root root 1173 may 11 2001 MAKEDEV
-rw-r--r-- 1 root root 11820 mar 21 18:55 Makefile
drwxr-xr-x 5 root root 2048 jun 20 00:33 nbd
drwxr-xr-x 2 root root 1024 nov 6 2002 patches
-rw-r--r-- 1 root root 1692 nov 21 2002 README
drwxr-xr-x 6 root root 1024 nov 8 2002 ssl
------------------------------------------------------------------------=
----
--------------
- I change the SMP value in Makefile to 1 because it=B4s a dual
proccesor
machine.
[root@nibble nbd-2.4.30]# uname -a
Linux nibble 2.4.9-e.3smp #1 SMP Fri May 3 16:48:54 EDT 2002 i686 =
unknown
------------------------------------------------------------------------=
----
--------------
Make clean...
[root@nibble nbd-2.4.30]# make clean
make -C /tmp VPATH=3D/mnt/data/tmp/nbd/nbd-2.4.30/nbd clean
make[1]: Cambiando a directorio `/tmp'
for x in enbd-server.o socket.o logging.o alarm.o file.o select.o
fileserver.o select.o db.o stream.o ioctl.o ipaddrfile.o pidfile.o
interface.o lockfile.o md5.o lock.o shmem.o enbd-client.o socket.o =
logging.o
alarm.o stream.o select.o ioctl.o netserver.o pidfile.o md5.o =
enbd-sstatd.o
config.o ipaddrfile.o lockfile.o select.o enbd-cstatd.o config.o =
interface.o
pidfile.o select.o enbd-test.o enbd-server enbd-sstatd enbd-client
enbd-cstatd enbd-test dummy; do rm -f /tmp/$x; done
for d in dummy; do \
test -d /mnt/data/tmp/nbd/nbd-2.4.30/nbd/$d && make -f
/mnt/data/tmp/nbd/nbd-2.4.30/nbd/$d/Makefile clean; \
done
make[1]: [clean] Error 1 (no tiene efecto)
make[1]: Saliendo directorio `/tmp'
cd /tmp && rm -f enbd.o .enbd.o.flags enbd_*.o .enbd_*.o.flags
[root@nibble nbd-2.4.30]#
------------------------------------------------------------------------=
----
--------------
Make config ...
root@nibble nbd-2.4.30]# make config
export CONFIG_SITE=3D/mnt/data/tmp/nbd/nbd-2.4.30/conf/config.Linux; \
cd /tmp; ./configure --srcdir=3D/mnt/data/tmp/nbd/nbd-2.4.30/nbd; \
make VPATH=3D/mnt/data/tmp/nbd/nbd-2.4.30/nbd
CFLAGS=3D"-D_LARGEFILE64_SOURCE=3D1 -D_LARGEFILE_SOURCE=3D1 =
-D_GNU_SOURCE=3D1 -D_XOP
EN_SOURCE=3D1 -D_FILE_OFFSET_BITS=3D64 -Wall -O2 -I/tmp =
-I/mnt/data/tmp/nbd/nbd-
2.4.30/linux/include -DDEBUG=3D0" EXTRA_LIBS=3D" " config
loading cache ./config.cache
checking for gcc... (cached) gcc
checking whether the C compiler
(gcc -D_LARGEFILE64_SOURCE=3D1 -D_LARGEFILE_SOURCE=3D1 =
-D_GNU_SOURCE=3D1 -D_XOPEN_
SOURCE=3D1 -D_FILE_OFFSET_BITS=3D64 -Wall -O2 ) works... yes
checking whether the C compiler
(gcc -D_LARGEFILE64_SOURCE=3D1 -D_LARGEFILE_SOURCE=3D1 =
-D_GNU_SOURCE=3D1 -D_XOPEN_
SOURCE=3D1 -D_FILE_OFFSET_BITS=3D64 -Wall -O2 ) is a cross-compiler... =
no
checking whether we are using GNU C... (cached) yes
checking whether gcc accepts -g... (cached) yes
checking whether warnings should be enabled... yes
checking for install... (cached) /usr/bin/install
checking whether byte ordering is bigendian... (cached) no
checking for inline... (cached) inline
checking for working const... (cached) yes
checking size of unsigned char... (cached) 1
checking size of unsigned short int... (cached) 2
checking size of unsigned int... (cached) 4
checking size of unsigned long int... (cached) 4
checking size of unsigned long long int... (cached) 8
checking size of char... (cached) 1
checking size of short int... (cached) 2
checking size of int... (cached) 4
checking size of long int... (cached) 4
checking size of long long int... (cached) 8
checking for llseek... (cached) yes
checking for lseek64... (cached) yes
checking for if_nameindex... (cached) yes
checking for bdflush... (cached) yes
checking for mlockall... (cached) yes
checking for strgsub in -lpub... (cached) no
checking how to run the C preprocessor... (cached) gcc -E
checking for ANSI C header files... (cached) yes
checking for size_t... (cached) yes
checking for socklen_t... yes
checking whether client should be built... yes
checking where your init scripts are kept... /etc/init.d
checking where pidfiles and state files go... below /var
checking where helper scripts go... ${exec_prefix}/sbin
checking where configuration files go... /etc
checking where man pages go... ${prefix}/man
checking where misc documents go... ${prefix}/share/doc
checking where daemons go... ${exec_prefix}/sbin
creating ./config.status
creating Makefile
creating config.h
config.h is unchanged
make[1]: Cambiando a directorio `/tmp'
make[1]: No se hace nada para `config'.
make[1]: Saliendo directorio `/tmp'
Make ...
(I've attached the output, because its easier to see).
I think that=B4s all.
Thank you very much.
PD: Sorry if it=B4s too long but I didn=B4t want to forget anything ;)
-----Mensaje original-----
De: enbd-admin@lists.community.tummy.com
[mailto:enbd-admin@lists.community.tummy.com]En nombre de Peter T.
Breuer
Enviado el: jueves, 19 de junio de 2003 3:09
Para: enbd@lists.community.tummy.com
Asunto: Re: [ENBD] (no subject)
"A month of sundays ago Javier Llorente wrote:"
> I am new with ENB and I need some help to compile the latest stable
version
> of enbd. I_m actually implementing RAC Oracle 9i and I am running it =
on
Red
> Hat Advanced Server 2.1, kernel 2.4.9 with no pacthes applied. I am =
not
sure
2.4.9 is too old for me to know if current versions of enbd compile
against it or not. It's pre- the great VMS changeover.
> if its possible, but I would like to use an stable version of enbd =
without
> pacthing any libraries of RH because it certified by Oracle and OCFS
(Oracle
Libraries would not be involved!
You'll have to tell me if it compiles or not. I'll make changes for you =
if
modifications are needed.
Peter
_______________________________________________
ENBD mailing list
ENBD@lists.community.tummy.com
http://lists.community.tummy.com/mailman/listinfo/enbd