Ufsdump and zones solaris 10

We do a backup daily via a script of the our zone using ufsdump.Sometimes the backup goes through each file system and other times the entire process is repeated three times.
I have also checked that there is no instruction entry specified in the script to try again if the backup should fail for some reason. Any help will do.

It depends on what file systems you are backing up. Perhaps you have a backup from the root file system on down, which would include the zones. Or, say for example you use /export/home for your zones and you backup /export/home on a regular basis; that job would also include the zones underneath.
ufsdump is not zone-aware in the sense that it will read zonepath settings and treat them as separate filesystems. It might simply be that some of your jobs are redundant.

Similar Messages

Patchadd and zones??

I feel dumb asking this question especially since I think its due to my lack of knowledge with zones. I am using Solaris 10 with only a global zone, at least thats what I think? I ran a "zoneadm list" and only listed "global".
I have downloaded uwc [yes I posted something similar on the JMS forum] patch 118540-42 and have tried to add it with patchadd. I get an error saying " Package SUNWuwc from directory SUNWuwc in patch 118540-42 is not installed on the system" and the patch does not install [I can not see it with showrev -p nor in /var/sadm/patch]. I also tried the "-G" option but no difference.
Then I tried using Solaris Management Console 2.1 using its patch tool. Action -> add patch and I get an error saying "Some or all of the software packages that patch 118540-42 patches are not installed on the target host.".
So I decided to look for SUNWuwc with pgkinfo and there is no listing, in fact none of the Java comm suite is found [nor most of the stuff in /opt and /usr/local ]. Yet I know JES is all in /opt and is working fine. I have never created a zone on this system and thought I had installed everything to just a global zone. I have had this system running for over a year and some updates have happened [just not sure what changed].
So I am wondering is there another way to add patches with Sol 10? another flag or utility? Is there a way to know what zone one is working in? or has installed stuff to. Is there a way at this point to stop using zones altogether?
Maybe I am missing a Sol 10 patch?
I had also run into a file /var/sadm/install/gz-only-packages and sure enough all my installed packages were listed here [all the JES stuff and web server etc]. I thought this meant "global zone only" and I only have a global zone??? so why can I not patch??
Apologies for the confusion [and lack of more details], although I have worked with Solaris for years this new 10 version and zones is a bit confusing for me. Plus the problem might not be zones related at all :^)
Thanks in Advance
-James

try zoneadm list -vc
this shows all zones on the system, whether they are configured, running or not running.
Next, try to get the 10-recommended patches from sunsolve.sun.com and add those, but do it in single-user mode.
/tony

IPMP configuration and zones - how to?

Hello all,
So, I've been thrown in at the deep end and have been given a brand new M4000 to get configured to host two zones. I have little zone experience and my last Solaris exposure was 7 !
Anyway, enough of the woe, this M4000 has two quad port NICs, and so, I'm going to configure two ports per subnet using IPMP and on top of the IPMP link, I will configure two v4 addresses and give one to one zone and one to the other.
My question is, how can this be best accomplished with regards to giving each zone a different address on the IPMP link.
IP addresses available = 10.221.91.2 (for zone1) and (10.221.91.3 for zone2)
So far, in the global zone I have
ipadm create-ip net2 <-----port 0 of NIC1
ipadm create-ip net6 <-----port 0 of NIC2
ipadm create-ipmp -i net2,net6 ipmp0
ipadm create-addr -T static -a 10.221.91.2/24 ipmp0/zone1
ipadm create-addr -T static -a 10.221.91.3/24 ipmp0/zone2
the output of ipmpstat -i and ipmpstat -a is all good. I can ping the addresses from external hosts.
So, how now to assign each address to the correct host. I assume I'm using shared-ip?
in the zonecfg, do I simply (as per [this documentation|http://docs.oracle.com/cd/E23824_01/html/821-1460/z.admin.task-54.html#z.admin.task-60] ):
zonecfg:zone1> add net
zonecfg:zone1:net> set address=10.221.91.2
zonecfg:zone1:net> set physical=net2
zonecfg:zone1:net> end
and what if I have many many addresses to configure per interface... for example zone1 and zone2 will also require 6 addresses on another subnet (221.206.29.0)... so how would that look in the zonecfg?
Is IPMP the correct way to be doing this? The client wants resilience above all, but these network connections are coming out of different switches thus LACP/Trunking is probably out of the question.
Many thanks for your thoughts... please let me know if you want more information
Solaris11 is a different beast altogether.
Edited by: 913229 on 08-Feb-2012 08:03
added link to the Solaris IPMP and zones doc

Thanks for the reply....
It still didn't work... but you pointed me in the right direction. I had to remove the addresses I had configured on ipmp0 and instead put them in the zonecfg. Makes sense really. Below I have detailed my steps as per your recommendation...
I had configured the zone as minimally as I could:
zonepath=/zones/zone1
ip-type=shared
net:
address: 10.221.91.2
physical=ipmp0
but after it is installed, I try and boot it and I get:
zone 'zone1': ipmp0:2: could not bring network interface up: address in use by zone 'global: Cannot assign the requested address
So, I changed the ip-type to exclusive and I got:
WARNING: skipping network interface 'ipmp0' which is used in the global zone.
zone 'zone1': failed to add network device
which was a bit of a shame.
So, finally, I removed the addresses from ipmp0
ipadm delete-addr ipmp0/zone1
ipadm delete-addr ipmp0/zone2
and set the address in zonecfg together with the physical=ipmp0 as per your recommendation and it seems to be working.
So, am I correct in taking away from this that if using IPMP in shared-ip zones, don't set the address in the global zone, but stick it in the zone config and everyone is happy?
I think this was the only way to achieve multiple IP addresses on one interface but over two ports?
Lastly, why oh why is the gateway address in netstat -rn coming up as the address of the host?
Anyway, thanks for your help.
;)

Default gateways and zones in a multihomed system

We do have some problems concerning default routes and zones in a multihomed system.
I found several posts in this forum, most of them referring to a domument of meljr, but my feeling ist that the paper is either not correct or not applicable to our situation?! Perhaps somebody can give me a hint.
Let me sketch our test environment. We have a multihomed Solaris 10 system attached to three different DMZ's using three different network adapters. We set up two local zones with IP's of the DMZ's of adapter 1 and 2, leaving adapter 0 for the IP of the global zone.
Now we set up default routes to ensure that network traffic from the local zones is routed in the corresponding DMZ's. That makes three different default routes on the global zone. On startup of the local zones, netstat reports the expected default routes to the correct DMZ gateways inside each zone.
Now what happens... My ssh to the global zone sometimes breaks. When this happens, no pings are possible to the IP of the global zone. Meanwhile, pings from other machines in our network (even from different subnets) might produce replies, some don't. By now, I can't tell you if there's is anything deterministic about it... More interesting: the local zone connections aren't affected at all!
So we did some more testing. Binding an IP address to the DMZ interfaces where the zones are tied to makes no difference (we tried both, with or without dedicated addresses for the adapter in the global zone). So the setup we're using right now is made of 5 IP addresses.
IP1, subnet 1: adapter 0, global zone
IP2, subnet 2: adapter 1, global zone
IP3, subnet 2; adapter 1. local zone 1
IP4, subnet 3; adapter 2, global zone
IP5, subnet 3; adapter 2, local zone 2
In the global zone there are three default gateways defined, one in each DMZ subnet. Inside the local zones, at startup you'll find the corresponding gateway into the DMZ. Everything looks fine...
I opened five ssh connections to the different IP's. Now what happened... After approx. half an hour, the connections to two IPs of the global zone (adapter 0 and adapter 1) broke down, while the connections to all other IP's were still open. This behaviour can be reconstructed!
So perhaps anybody has a explaination for this behaviour. Or perhaps anybody can answer me some qustions:
1. How are the three default gateways handled? Is there still some kind of "round robbin" implemenation? How can I guarantee that network traffic from outside isn't routed inside the DMZ's without preventing the local zones from talking to each other (actually we only need to communicate on some ports, but the single IP-stack concept only gives us all or nothing...).
2. If I do a ping from local zone 1 to the default gateway of local zone 2, this route is added as additional default gateway inside local zone 1! So does this mean, the routing decision is made only inside the global zone not taking into account where the packet is sent from?
3. After all, how are the IP packets routed from the different zone and the global zone, and how are they routed back to calling systems from the various DMS's and other networks, routed via these DMS's???
The scenario seems to be covered by http://meljr.com/~meljr/Solaris10LocalZoneDefaultRoute.html, but configuring the machine like stated in the paper leaves me with the problems described.
I'd be happy for any helpful comment!

you can have multiple gateway entries in deafultrouter file but the default gateway for global zone can be only one but you can specify different gateways for different zones..
using this default gateway, you should be able to connect via different network...!

How to wipe the old Sun OS and install Solaris 10?

hello,
I bought a Sun Ultra 5 server from ebay, and wanted to install Solaris 10 on it. I downloaded the 5-CD version of Solaris 10 from Sun website.
When I powered on the server, it dispalys a series of prompts:
Type control-d to proceed with normal starup
(or give root password for system maintenance)
I typed control-d a couple of times, and eventually it asked me to select run level 0-6, s
no matter which one I chose, it went back to 'type control-d ..' prompt.
How am I supposed to start the installation program? (I have CD 1 in the CD-ROM)
Or, how to completely wipe the old system and install Solaris 10?
many thanks

If you get to select a run lever try 0 (halt) that will bring you to an ok> prompt where you can boot cdrom.
Alternately try STOP-A from the keyboard. That should also drop you to an ok> prompt and you can boot cdrom from there.

EBS 7.4 with ZFS and Zones

The EBS 7.4 product claims to support ZFS and zones, yet it fails to explain how to recover systems running this type of configuration.
Has anyone out there been able to recover a server using the EBS software, that is running with ZFS file systems both in the Global zone and sub zones (NB: The servers system file store / /usr /var, is UFS for all zones).
Edited by: neilnewman on Apr 3, 2008 6:42 AM

The EBS 7.4 product claims to support ZFS and zones, yet it fails to explain how to recover systems running this type of configuration.
Has anyone out there been able to recover a server using the EBS software, that is running with ZFS file systems both in the Global zone and sub zones (NB: The servers system file store / /usr /var, is UFS for all zones).
Edited by: neilnewman on Apr 3, 2008 6:42 AM

Minimize and maximize Solaris 10 installed on VM virtual Box.

How can I minimize and maximize Solaris 10 installed on VM virtual Box.
Thanks
Paulo

Solaris is working just fine. Everything is ok. now since I open Solaris via VM Virtual box, the Solaris main page fills only half of the monitor. I can minimize/maximize the Virtual Box. But Solaris dimensions are fix and doesn't change. This leaves me a little room to open applications on Solaris. I had the same problems with my Ubuntu, by installing adds-on it was fixed. For Solaris I installed Guest additions, didnt help.
Thanks
Paulo

Combining ufsdump and a tape library

Hi All,
I have a question with regards to multi-volume backups using ufsdump and a tape library. Reading through the ufsdump documentation, I see that it has an option "l" (lowercase L), which says "When the end-of-tape is reached before the dump is complete, take the drive offline and wait up to two minutes for the tape drive to be ready again ...". However, when I use this option, and ufsdump reaches the end of the tape, it does not offline the tape drive. It sits there, waiting for a response to the question "DUMP: NEEDS ATTENTION: Do you want to restart?: ("yes" or "no")"...at no time does the tape drive go offline, which would allow me to write a script to detect the offlining of the tape, and swap it out for a fresh one.
Basically, what i'm looking for is a way to allow ufsdump to span >1 tapes without operator intervention by having a script detect when it needs new media, and ask the library to load the next tape. However, ufsdump doesn't relinquish control of the tape drive, so I can't pop out the tape unless I kill it...
Can anyone suggest a solution to automating multi-volume ufsdumps to tape?
Thanks,
-Adam Lundrigan
[email protected]

I don't think you can do this AND have a variable sized output media. You will get the restart question when it gets a tape error (which you will get if you write out too much). I think the only way you can do what you want is to conservatively figure out what would fit on a tape with compression, and giving a tape length where the length*bpi will give you that figure (ie, no errors on tape writes).
-r

Ufsdump and ufsrestore with multiple volumes

How to ufsdump and ufsrestore with multiple volumes?
I want to backup and restore a large directory with ufsdump and
ufsrestore, so it requires multiple tapes (two in this example).
After backuping tape 1, I press the eject button of the tape drive to
eject the button, then I insert a second tape and type "yes" to
continue.
# /usr/sbin/ufsdump 0uf /dev/rmt/0cn /opt/bsm
DUMP: 89.66% done, finished in 0:04
DUMP: End-of-tape detected
DUMP: 92.68% done, finished in 0:03
DUMP: Change Volumes: Mount volume `#2' on `pwdly0ee:/dev/rmt/0cn'
DUMP: NEEDS ATTENTION: Is the new volume (#2) mounted on
`pwdly0ee:/dev/rmt/0cn' and ready to go?: ("yes" or "no")
DUMP: NEEDS ATTENTION: Is the new volume (#2) mounted on
`pwdly0ee:/dev/rmt/0cn' and ready to go?: ("yes" or "no")
DUMP: NEEDS ATTENTION: Is the new volume (#2) mounted on
`pwdly0ee:/dev/rmt/0cn' and ready to go?: ("yes" or "no") ye
DUMP: NEEDS ATTENTION: Is the new volume (#2) mounted on
`pwdly0ee:/dev/rmt/0cn' and ready to go?: ("yes" or "no") yes
DUMP: "yes" or "no"?
DUMP: NEEDS ATTENTION: Is the new volume (#2) mounted on
`pwdly0ee:/dev/rmt/0cn' and ready to go?: ("yes" or "no") yes
DUMP: Volume 2 begins with blocks from inode 430787
DUMP: 92.70% done, finished in 0:03
DUMP: Change Volumes: Mount volume `#2' on `pwdly0ee:/dev/rmt/0cn'
DUMP: NEEDS ATTENTION: Is the new volume (#2) mounted on
`pwdly0ee:/dev/rmt/0cn' and ready to go?: ("yes" or "no") yes
DUMP: Volume 2 begins with blocks from inode 430787
DUMP: 92.70% done, finished in 0:03
DUMP: 4731838 blocks (2310.47MB) on 2 volumes at 840 KB/sec
DUMP: DUMP IS DONE
DUMP: Level 0 dump on Sat 08 Nov 2003 02:53:58 PM EST
# /usr/sbin/ufsrestore rf /dev/rmt/0cn
Mount volume 2
then enter volume name (default: /dev/rmt/0cn)
Mount volume 3
then enter volume name (default: /dev/rmt/0cn) <- insert second tape
Wrong volume (2)
Mount volume 3
then enter volume name (default: /dev/rmt/0cn) 2
Cannot open 2
Mount volume 3
then enter volume name (default: 2)
Cannot open 2
Mount volume 3
After tape 1 in finished restoring, I eject the tape and insert the
second tape.
Now what should I do to continue tape restoring?

I don't think that you're supposed to type in the number "2". What ufsrestore is looking for is the next ufsrestore volume, not the volume number. If it is a tape drive, which this apears to be, you specify the tape drive device: dev/rmt/0cn.

SunMC 3.6.1 Agent and zones

Hi folks,
I'm wondering how to handle zones with SunMC 3.6.1. We're planning to use zones in the future, mostly whole root zones. As far as I understood, I need to install the agent in the global zone and in the local zone afterwards. Is this correct?
Thanks in advance

1. How does the container manager link with SunMC.
Does monitor info go to the sunMC server ?SCM is one of the optional components you're prompted to add when you're installing SunMC Server and Agents. The way it works on the Server is that SCM adds an extra web interface, and it relies on the SunMC for host lists and a place to store it's data (i.e the graphs shown in the SCM web interface are actually coming from the SunMC Oracle database).
On the Agent side, all that happens when you say "y"es to SCM during install is that it adds a hidden module to the Agent that collects container/zone-type data, and can do the bidding of the SCM web interface (i.e. make/destroy Zones, alter CPU pools and container shares etc)
Do you only have to install the container manager
on the global zone ?Yes, it's only required for the SunMC Agent in each global zone. It's that Agent that can "see everything" about the containers and zones that are running.
Regards,
[email protected]
http://www.HalcyonInc.com

Saposcol, Solaris 10 and zones

We are running a bunch of instances on Solaris 10 x86_64 in zones. Everything works great (including ZFS) but one issue is there:
saposcol always "sees" the full machine, means e. g. CPU usage is displayed not on a per-zone basis but only once for all zones.
Are there any plans on extending saposcol to support zone-specific data (such as IBM has recently done with AIX-micropartitioning and/or LPARS)? It would greatly improve resource management on those systems.
Markus

Yup i do see em at unix lvl.
Filesystem             size   used avail capacity Mounted on
/                       20G   6.8G    13G    35%    /
/archivos              1.9G   2.0M   1.9G     1%    /archivos
/dev                    20G   6.8G    13G    35%    /dev
/dvds                   39G   6.6G    32G    17%    /dvds
/sapdb/PRD/db          962M   167M   737M    19%    /sapdb/PRD/db
/sapdb/PRD/logbackups
                       4.9G   5.0M   4.9G     1%    /sapdb/PRD/logbackups
/sapdb/data            1.9G    62M   1.8G     4%    /sapdb/data
/sapdb/programs        962M   208M   696M    24%    /sapdb/programs
/sapmnt/PRD            1.9G   686M   1.2G    36%    /sapmnt/PRD
/usr/sap/PRD           7.9G   754M   7.1G    10%    /usr/sap/PRD
/usr/sap/trans         1.9G   414M   1.5G    22%    /usr/sap/trans
proc                     0K     0K     0K     0%    /proc
ctfs                     0K     0K     0K     0%    /system/contract
swap                    31G   296K    31G     1%    /etc/svc/volatile
mnttab                   0K     0K     0K     0%    /etc/mnttab
/platform/sun4u-us3/lib/libc_psr/libc_psr_hwcap2.so.1
                        20G   6.8G    13G    35%    /platform/sun4u-us3/lib/libc_psr.so.1
/platform/sun4u-us3/lib/sparcv9/libc_psr/libc_psr_hwcap2.so.1
                        20G   6.8G    13G    35%    /platform/sun4u-us3/lib/sparcv9/libc_psr.so.1
fd                       0K     0K     0K     0%    /dev/fd
swap                    31G     0K    31G     0%    /tmp
swap                    31G    56K    31G     1%    /var/run
and saposcol is running with root user.
Message was edited by:
        Daniel Esteban Rajmanovich

Solaris 10 6/06 ZFS and Zones, not quite there yet...

I was all excited to get ZFS working in our environment, but alas, a warning appeared in the docs, which drained my excitement! :
http://docs.sun.com/app/docs/doc/817-1592/6mhahuous?a=view
essentially it says that ZFS should not be used for non-global zone root file systems.. I was hoping to do this, and make it easy, global zone root UFS, and another disk all ZFS where all non-global whole root zones would live.
One can only do so much with only 4 drives that require mirroring! (x4200's, not utilizing an array)
Sigh.. Maybe in the next release (I'll assume ZFS will be supported to be 'bootable' by then...
Dave

I was all excited to get ZFS working in our
environment, but alas, a warning appeared in the
docs, which drained my excitement! :
http://docs.sun.com/app/docs/doc/817-1592/6mhahuous?a=
view
essentially it says that ZFS should not be used for
non-global zone root file systems..Yes. If you can live with the warning it gives (you may not be able to upgrade the system), then you can do it. The problem is that the the installer packages (which get run during an upgrade) don't currently handle ZFS.
Sigh.. Maybe in the next release (I'll assume ZFS
will be supported to be 'bootable' by then...Certainly one of the items needed for bootable ZFS is awareness in the installer. So yes it should be fixed by the time full support for ZFS root filesystems is released. However last I heard, full root ZFS support was being targeted for update 4, not update 3.
Darren

Reinstall zones (solaris 8 container)

Hi
I have to restore 2 zones, because originally the zones are stored on a ZFS filesystem i've did the backup with the zfsdump scripts.
I'm able to restore the ZFS filesystems but that leaves me with the problem of reconfiguring the zones.
how do i configure and reinstall these zones?, the original flash archive used to setup these zones is missing.
would it be possible to use the following procedure
1) restore the zfs filesystemen
2) clone the latest snapshot and mount it.
3) create an archive from the mounted snapshot to use as the install media for the zone
the second option i can think of is
1 ) create a standard solaris 8 zone with the solaris 8 image you can download from Sun
2) Copy the files from the mounted clone to the new zone
3) boot the zone
I'm hoping someone could guide me in the right direction
Regards,

I just managed to restore a zfs filesystem with a solaris 8 container on it using the steps below.
I'm curious if i got it right or if there is another more obvious way as i have only a little experience and hands on time with the zfs filesystem.
bash-3.00# zfsrestored.sh -z storage/z-chili
storage/z-chili is restored from /nsr/storage_z-chili.20081206.
bash-3.00#
bash-3.00# zfs clone storage/z-chili@20081206 storage/z-chili/restored
bash-3.00# zfs list
NAME USED AVAIL REFER MOUNTPOINT
storage 85.9G 48.0G 24.5K legacy
storage/data 69.7G 48.0G 69.7G /data
storage/z-chili 7.61G 48.0G 7.61G legacy
storage/z-chili@20081206 0 - 7.61G -
storage/z-chili/restored 0 48.0G 7.61G legacy
storage/z-polen 4.26G 48.0G 4.26G legacy
storage/z-polen@20081211 832K - 4.26G -
storage/zones 4.32G 48.0G 4.32G /export/zones/
bash-3.00#
zfs set mountpoint=/mnt/restore storage/z-chili/restored
bash-3.00# zfs list
NAME USED AVAIL REFER MOUNTPOINT
storage 85.9G 48.0G 24.5K legacy
storage/data 69.7G 48.0G 69.7G /data
storage/z-chili 7.61G 48.0G 7.61G legacy
storage/z-chili@20081206 0 - 7.61G -
storage/z-chili/restored 0 48.0G 7.61G /mnt/restore
storage/z-polen 4.26G 48.0G 4.26G legacy
storage/z-polen@20081211 832K - 4.26G -
storage/zones 4.32G 48.0G 4.32G /export/zones/
bash-3.00# flar create -n zchili -R /mnt/restore/root/ /data/z-chili-restored.flar
Full Flash
Checking integrity...
Integrity OK.
Running precreation scripts...
Precreation scripts done.
Determining the size of the archive...
and i've used the flar image to install the zone.
apart from some file/dir permission rights this worked and i got the zone and webservers up and running.

Problem migrate zones solaris 10

Hi,
i'm trying to migrate one non-global zone from one machine to another machine. The steps I have followed are:
1) On the machine where I have the zone that I want to migrate:
# zonecfg -z myzone export -f /myzone.cfg
# cd /export
# find zones/myzone -fstype lofs -prune -o -fstype ufs | cpio -ocm -O /myzone.cpio
2) On the second machine, I have copied the myzone.cfg and myzone.cpio, and after:
- I modified in myzone.cfg the option set physical with the correct one for this machine
# zonecfg -z myzone -f /myzone.cfg
# cd /export
# cat /myzone.cpio | cpio -icdm
# cp /etc/zones/index /etc/zones/index.bak
# perl -pi -e 's/^myzone:configured/myzone:installed/' /etc/zones/index
[root@pontevedra]# zoneadm list -icv
ID NAME STATUS PATH
0 global running /
- myzone installed /export/home/zones/myzone
But I have a problem when I boot the zone:
[root@pontevedra]# zoneadm -z myzone boot
zoneadm: zone 'servicio-zone': mount of /proc failed: No such file or directory
zoneadm: zone 'servicio-zone': call to zoneadmd failed
I don't know what happen... can anybody helpme?
Thanks a lot in advance!

Hi Darren,
Thanks God I've succesfully installed Solaris 10 to my laptop, after "swimming" for a while in google :)
It said that this was a bug of Solaris 10 when installed in PC with Pentium III or II processors.
Here is the solutions (hope someone with same problem with me could find it useful) :
1. After the GRUB menu appears, edit the kernel line then append -kd at the end (to enter kmdb upon boot), then boot
2. Then type this command when kmdb prompt shows, [0]>
[0]> cmi_no_init/W1
[0]> :c 3. It should boot and run without problem, then follow the installation menu and option
4. After reboot, again follow step 1 & 2 when GRUB menu shows. I promise this is the last time :)
5. To make it permanent, add this following instruction to */etc/system* once the system up
set cmi_no_init=1//Henry

Geographic Cluster and zones

Hello,
I am attempting to setup a geographic cluster to failover an Informix application to our disaster recovery (BCP) site. I have a Sun Fire V440 in each location running Solaris 10 08/07 update. The application is currently running on a Solaris 8 02/04 server and must continue to do so. The catch is that the server in the BCP site is also used as a QA server. My thought was to create on the Solaris 10 server, two Solaris 8 containers, one for failover from the home office and the other used for QA. At the home office site, the server would run one Solaris 8 container. We are using EMC SRDF for replication and storage of the Informix database. The container on the home office server would failover to the BCP container on the server in the BCP site. My questions are: 1) Is this scenario possible, and 2) How would I configure the clustering on the servers? Should I be using the the data services for Containers and for Informix? I have so far created one node clusters in each site and was in the process of configuring the resource groups but was unsure how to proceed. Thanks for any help anyone can give. Thanks.

I work for the Sun Cluster Geographic Edition (SCGE) team so I hope I can give some definitive answers...
First, I'm slightly confused as to whether this is a single cluster with geographically split nodes or two single node clusters joined together with SCGE. From re-reading your posting it looks like it is the latter, which although far from optimal, is possible. The point to make is that any failure on the primary site is probably going to have to be treated as a disaster. You will need to decide whether the primary site node will be back up any time soon and if not, take-over the service on the remote (DR) site. Once you've taken over the service, the reverse re-sync is going to be quite expensive. If you'd had a local cluster at the primary site, then only site failures would have forced this decision.
Back to the configuration. You'll need to install single node Solaris Clusters (3.2 01/09) at each site. You would then create an HA-container agent resource group/resource to hold you Solaris 8 brand zone. You'd then put your Informix database in this container. You'd do the same at the remote site. Your options for storage are raw DIDs devices or file system. You can't use SVM with SRDF yet and I don't think there is a supported way to map VxVM volumes into the HA-container (though I may be wrong). Personally, I'd use UFS on raw DID (or VxVM) in the global zone mounted with forcedirectio and map that into the HA-container. (http://docs.sun.com/app/docs/doc/820-5025/ciagbcbg?l=en&a=view)
I don't know off-hand whether the the Informix agent will run in with an HA-container agent with a Solaris 8 brand container. I'll ask a colleague.
If you need any more information, it might be more helpful to contact me directly at Sun. (First.Last)
Regards,
Tim
---

Ufsdump and zones solaris 10

Similar Messages

Maybe you are looking for