Failover Zone Creation

Still at the beginning of the learning curve here....
I am looking to create a Solaris 10 failover zone in my 3.2 Cluster.
I see documents claiming both ways that the zone can or cannot have its zonepath on a globally shared file system.
So, do I
Create equivalent zones on the local drives, then add a pointer in the zone to a file system on the global area, then go forward with my cluster resource creation from there
-or-
No, just create the zone on the shared storage area once from node 1, the delete it and create it from node 2, then bring in HAStorageplus, etc.
Thanks.

I've been trying to configure a failover zone on my cluster and I'm having a terrible time. I do remember seeing somewhere that zones needed to be on a local filesystem, but I can't find that now.
I'm following these instructions: http://docs.sun.com/app/docs/doc/819-3059/6n5aafjor?a=view
Everything goes as expected and I have this:
$ clrg status
Cluster Resource Groups ===
Group Name Node Name Suspended Status
nfs-rg stnv240a No Online
stnv240b No Offline
web-rg stnv240a No Online
stnv240b No Online
zone-rg stnv240b No Online
stnv240a No Offline
$ clrs status
Cluster Resources ===
Resource Name Node Name State Status Message
cluster stnv240a Online Online - LogicalHostname online.
stnv240b Offline Offline
nfs-stor stnv240a Online Online
stnv240b Offline Offline
nfs-res stnv240a Online Online - Service is online.
stnv240b Offline Offline
web-stor stnv240a Online Online
stnv240b Online Online
zone-rs stnv240b Online Online
stnv240a Offline Offline
$ zonecfg -z cwm -f cwm-failover.cfg
$ zonecfg -z cwm info
zonename: cwm
zonepath: /global/nfs/zones/cwm
brand: native
autoboot: false
bootargs:
pool:
limitpriv:
scheduling-class:
ip-type: shared
net:
     address: 204.155.141.145
     physical: bge0
     defrouter not specified
But then I can't boot. I got a similar error trying to configure a sparse zone. Install fails there, as it can't mount /lib
$ zoneadm -z cwm boot
zoneadm: zone 'cwm': "/usr/lib/fs/lofs/mount -o zonedevfs /global/nfs/zones/cwm/dev /global/nfs/zones/cwm/root/dev" failed with exit code 33
zoneadm: zone 'cwm': call to zoneadmd failed
Any help or ideas here?
Solaris 10 10/08 s10s_u6wos_07b SPARC
Solaris Cluster 3.2 u1 but patched so I have sczonecluster.

Similar Messages

Will zones inherit security/hardening settings made prior to zone creation?

Hi guys.
So the scenario is I've just finished hardening Solaris 10: disabling of services, configuring tcp wrappers and the such. I would now like to create a zone or zones but there are a few things I would like to make sure of, before proceeding.
The questions are:
i) After creating a zone, whether it's sparse or whole, will the non-global zone inherit the hardening settings that I have setup prior? Or is it a seperate entity that requires me to harden it.
ii)I understand that patching the global zone will patch the rest of the other zones. Does this include whole-root zone since it does not mount any filesystem outside its zone.
Thank you.
N.

No, except for the packages that get installed at zone creation time, they're separate.
Patches are applied to all zones. A zone may be temporarily booted just to apply a patch. Even though a whole-root zone may not mount external filesystems, that doesn't mean that it's independent. For instance, it's kernel files aren't actually loaded, since it doesn't have an independent kernel. When the global zone's kernel is patched, all others are as well. So the files within the zone will be patched as well to maintain consistency.
Darren

Removing Packages "added" by zone creation

The search feature on these forums is limited. Scrolling back about 8 months I think I know the answer to this question, but I would like to be certain.
I have a solaris 10 global zone with SUNWut (sunray services) installed.
When I create my non-global zones, Solaris is "kind" enough to install the sunray services into the non-global zones as they get created. This leads to multiple instance of the sunray software running.
When I created the non global zones the first time, I went into the non--global zones and I removed the SUNWut packages using pkgrm.
Is there a way to tell the zone creation to not install the SUNWut package when the non global zone is created? I have to rebuild this machine and I would love to create the zones in a more efficient manner.
Looking through the posts I saw that people were dealing with my problem the same way I did. Once the non global zone is running, go in and do a pkgrm on it.
I just want to verify that this is the way other people are handling it.
Thanks,
Eric

The zones are a typical sparse root model (inherited-pkg-dir's are /lib/ /platform/ /sbin/ and /usr/). There is no nfs mounting or such.
Once the non-global zone is installed, If you go to /opt in the non-global zone the SUNWut directory is there. If you do a ps command you can see that there is a sunray server is running in each zone.
Thinking about it I suppose that the install script for SUNWut may be at fault. Perhaps it is telling the zone install that it is a base package when it is not. This would make sense because I was using a SRSS 3.0 install patched by my sun engineers to install on my sol10 system. Perhaps I will not run into the same issue now that I have the official 3.1 sunray s/w.
Only time will tell. But I appreciate the reply so that I know if it does happen that I am handling the problem in the correct manner (using pkgrm on SUNWut).
I will report back if this issue comes up again when I redo the system.
Thanks,
Eric

How to configure Oracle database in a failover zone cluster

Setup: Oracle database and zone configured on highly available local filesystems.
Two node cluster.
Oracle database running inside the zone.
Note: I dont have a zone cluster.
1. I need to make the zone and the oracle database highly available.
2. Can I configure the Oracle Data service directly to run in the zone or will it involve creating sczsh_config script to do the same.
I have been going through the guides and searching over the net, but haven't found any help so far. Its so much simpler to configure this environment in Veritas cluster. Hope I find some help here.
TIA,
Sudhir

The Oracle Solaris Cluster concepts guide has some information on which zone model to choose:
http://docs.oracle.com/cd/E18728_01/html/821-2682/gcbkf.html#scrolltoc
When managing a Solaris zone with the HA Zones agent, the cluster basically regards the non-global zone as a blackbox.
As such you can either start the Oracle database as part of the runlevel/SMF startup of the non-global zone, or you can use the sczsh or sczsmf component and use your own scripts to start, probe and stop the Oracle database.
Usage of the standard HA Oracle data service is not supported in combination with the HA Zones agent.
If you require a more fine grained control of services running in non-global zones, why not setup a zone cluster and then having HA Oracle failover the Oracle database between non-global zones?
Regards
Thorsten

Problem with zone creation in Solaris 10 6/06

Hi,
I have configured and installed a non-global zone following the example given in the guide http//www.sun.com/software/solaris/howtoguides/posgresqlhowto.jsp. My zone has the following configuration:
# zonecfg -z pg_zone
zonecfg:pg_zone> info
zonepath: /export/zones/pg_zone
autoboot: true
pool:
fs:
dir: /pg_log
special: /dev/dsk/c1d0s0
raw: /dev/rdsk/c1d0s0
type: ufs
options: []
net:
address: 192.168.10.100/24
physical: bge0
zonecfg:pg_zone> exit
The installation completed but I got the following errors:
# zoneadm -z pg_zone install
Preparing to install zone <pg_zone>.
Creating list of files to copy from the global zone.
Copying <130101> files to the zone.
Initializing zone product registry.
Determining zone package initialization order.
Preparing to initialize <1059> packages on the zone.
Initialized <1059> packages on zone.
Zone <pg_zone> is initialized.
Installation of these packages generated errors: <SUNWgnome-a11y-libs-share>
Installation of <63> packages was skipped.
The file </export/zones/pg_zone/root/var/sadm/system/logs/install_log> contains a log of the zone installation.
When examined the logs there were several messages like this:
SUNW_PKG_THISZONE=true in the pkginfo file.
*** package <SUNWamutl> was not installed:
The package <SUNWamutl> is registered to be installed in the
global zone only. The package may have been installed in the global
zone using the pkgadd '-G' option, or the package may contain a
request script, or the package may directly set the package attribute
SUNW_PKG_THISZONE=true in the pkginfo file.
*** package <SUNWiimdv> was not installed:
The package <SUNWiimdv> is either not fully installed in the
global zone, or the package contains a request script. Only packages
that are fully installed in the global zone and that do not contain a
request script can be installed when a new zone is created. If this
package does not contain a request script, you can correct this
problem by removing and reinstalling the package in the global zone.
And when I try to boot the zone I get this error:
# zoneadm -z pg_zone boot
zoneadm: zone 'pg_zone': "/usr/lib/fs/lofs/mount -o zonedevfs /export/zones/pg_zone/dev /export/zones/pg_zone/root/dev" failed with exit code 33
zoneadm: zone 'pg_zone': call to zoneadmd failed
Any suggestion how to solve this problem?
thanks

To answer your question wangzh: Yes I did. I actually followed exactly the procedure on the how-to-uides: http://www.sun.com/software/solaris/howtoguides/containersLowRes.jsp.
I think the problem here might reside in the script supplied in the solaris distribution for creating zones. For example, in the default configuration script, the package directories /lib /platform /var and /usr are by default set to be automatically imported. But when I try to create a zone using these defaults settings, it ends up with an mounting error that prevents the installation of the zone. only when I remove manually those directories from the configuration then the installation continues.
And during the installation process there are number of packages that are by default imported via the installation script and this is some of those packages that cause the errors here. I also tried to exclude the offending packages as suggested by henryC but I couldn't find all of them and the creation still fails at the booting.

Failover Zones / Containers with Sun Cluster Geographic Edition and AVS

Hi everyone,
Is the following solution supported/certified by Oracle/Sun? I did find some docs saying it is but cannot find concrete technical information yet...
* Two sites with a 2-node cluster in each site
* 2x Failover containers/zones that are part of the two protection groups (1x group for SAP, other group for 3rd party application)
* Sun Cluster 3.2 and Geographic Edition 3.2 with Availability Suite for SYNC/ASYNC replication over TCP/IP between the two sites
The Zones and their application need to be able to failover between the two sites.
Thanks!
Wim Olivier

Fritz,
Obviously, my colleagues and I, in the Geo Cluster group build and test Geo clusters all the time :-)
We have certainly built and tested Oracle (non-RAC) configurations on AVS. One issue you do have, unfortunately, is that of zones plus AVS (see my Blueprint for more details http://wikis.sun.com/display/BluePrints/Using+Solaris+Cluster+and+Sun+Cluster+Geographic+Edition). Consequently, you can't built the configuration you described. The alternative is to sacrifice zones for now and wait for the fixes to RG affinities (no idea on the schedule for this feature) or find another way to do this - probably hand crafted.
If you follow the OHAC pages (http://www.opensolaris.org/os/community/ha-clusters/) and look at the endorsed projects you'll see that there is a Script Based Plug-in on the way (for OHACGE) that I'm writing. So, if you are interested in playing with OHACGE source or the SCXGE binaries, you might see that appear at some point. Of course, these aren't supported solutions though.
Regards,
Tim
---

How to add network information for failover zones with logical hostname?

Hello!
As stated in [http://docs.sun.com/app/docs/doc/819-3069/ds_template-21?a=view] I must not configure network addresses for a zone when I manage these with a logical hostname:
If you require the SUNW.LogicalHostName resource type to manage all the zone's addresses, configure a SUNW.LogicalHostName resource with a list of the zone`s addresses and do not configure them by using the zonecfg utility.But when I start the zone for the first time using "zlogin -C" it does not ask me any questions about the network. Of course, there is no adapter configured. But how do I add information like routes or nameservers to the system when using a logical hostname?
TIA
Stephan

Hi Stephan,
I can only assume that when the zone was configured via zonecfg without any network interfaces that sysidcfg did not ask you for the default route or name service, as such you will need to setup those parts up manually.
Please take a look at the FAQs for zones, i.e. http://opensolaris.org/os/community/zones/faq/ in particular
http://opensolaris.org/os/community/zones/faq/#u5
http://opensolaris.org/os/community/zones/faq/#cfg_defroute
Finally, if you require a NIS client then please see http://docs.sun.com/app/docs/doc/816-5166/ypinit-1m?a=view
Regards
Neil

Failover Cluster creation fails 2012R2

Create Cluster
Cluster: FailoverCluster
Node: WS2012R2-2.yottabyte.inc
Node: WS2012R2-1.yottabyte.inc
IP Address: DHCP address on 192.168.136.0/24
Started 3/31/2015 11:48:52 PM
Completed 3/31/2015 11:52:13 PM
Beginning to configure the cluster FailoverCluster.
Initializing Cluster FailoverCluster.
Validating cluster state on node WS2012R2-2.yottabyte.inc.
Find a suitable domain controller for node WS2012R2-2.yottabyte.inc.
Searching the domain for computer object 'FailoverCluster'.
Bind to domain controller \\WS2012R2-1.yottabyte.inc.
Check whether the computer object FailoverCluster for node WS2012R2-2.yottabyte.inc exists in the domain. Domain controller \\WS2012R2-1.yottabyte.inc.
Computer object for node WS2012R2-2.yottabyte.inc does not exist in the domain.
Creating a new computer account (object) for 'FailoverCluster' in the domain.
Check whether the computer object WS2012R2-2 for node WS2012R2-2.yottabyte.inc exists in the domain. Domain controller \\WS2012R2-1.yottabyte.inc.
Creating computer object in organizational unit CN=Computers,DC=yottabyte,DC=inc where node WS2012R2-2.yottabyte.inc exists.
Create computer object FailoverCluster on domain controller \\WS2012R2-1.yottabyte.inc in organizational unit CN=Computers,DC=yottabyte,DC=inc.
Check whether the computer object FailoverCluster for node WS2012R2-2.yottabyte.inc exists in the domain. Domain controller \\WS2012R2-1.yottabyte.inc.
Configuring computer object 'FailoverCluster in organizational unit CN=Computers,DC=yottabyte,DC=inc' as cluster name object.
Get GUID of computer object with FQDN: CN=FAILOVERCLUSTER,CN=Computers,DC=yottabyte,DC=inc
Validating installation of the Network FT Driver on node WS2012R2-2.yottabyte.inc.
Validating installation of the Cluster Disk Driver on node WS2012R2-2.yottabyte.inc.
Configuring Cluster Service on node WS2012R2-2.yottabyte.inc.
Validating installation of the Network FT Driver on node WS2012R2-1.yottabyte.inc.
Validating installation of the Cluster Disk Driver on node WS2012R2-1.yottabyte.inc.
Configuring Cluster Service on node WS2012R2-1.yottabyte.inc.
Waiting for notification that Cluster service on node WS2012R2-2.yottabyte.inc has started.
Forming cluster 'FailoverCluster'.
Unable to successfully cleanup.
An error occurred while creating the cluster and the nodes will be cleaned up. Please wait...
An error occurred while creating the cluster and the nodes will be cleaned up. Please wait...
There was an error cleaning up the cluster nodes. Use Clear-ClusterNode to manually clean up the nodes.
There was an error cleaning up the cluster nodes. Use Clear-ClusterNode to manually clean up the nodes.
An error occurred while creating the cluster.
An error occurred creating cluster 'FailoverCluster'.
This operation returned because the timeout period expired
To troubleshoot cluster creation problems, run the Validate a Configuration wizard on the servers you want to cluster.
While creating cluster I am getting this error? Any solution please.
Thank You

Did you follow the instruction "To troubleshoot cluster creation problems, run the Validate a Configuration wizard on the servers you want to cluster"? Any warnings/errors?
. : | : . : | : . tim

Solaris10 branded zone creation error on Solaris 11.1

Hello,
We are having the problem in creating the Solaris10 branded zone on the solaris 11.1 version. We are trying to install this branded zone using the solaris10 VM template. Please can some one through some pointers ?
The installation steps stops at "Waiting for boot to complete" but never successfully boots even waiting for long times. We can see that Zone status in configured only.
root@localhost:~# /home/solaris-10u10-sparc.bin -p /solaris10 -i net0 -a xxx.xxx.xxx.xxx -z zone10
This is an Oracle VM Template for Oracle Solaris Zones.
Copyright Â© 2011, Oracle and/or its affiliates. All rights reserved.
This software and related documentation are provided under a license
agreement containing restrictions on use and disclosure and are
protected by intellectual property laws. Except as expressly
permitted in your license agreement or allowed by law, you may not
use, copy, reproduce, translate, broadcast, modify, license,
transmit, distribute, exhibit, perform, publish, or display any
part, in any form, or by any means. Reverse engineering,
disassembly, or decompilation of this software, unless required by
law for interoperability, is prohibited.
Checking disk-space for extraction
Ok
Extracting in /root/bootimage.dnaqaD ...
100% [===============================>]
Checking data integrity
Ok
Checking platform compatibility
The host and the image do not have the same Solaris release:
host Solaris release: 5.11
image Solaris release: 5.10
Will create a Solaris 10 branded zone.
IMAGE: /home/solaris-10u10-sparc.bin
ZONE: zone10
ZONEPATH: /solaris10/zone10
INTERFACE: net0
VNIC: vnicZBI6783
MAC ADDR: 2:8:20:5c:92:f5
IP ADDR: xxx.xxx.xxx.xxx
NETMASK: 255.0.0.0
DEFROUTER: # # This file is deprecated. Default routes will be created for any router # addresses specified here, but they will not change when the underlying # network configuration profile (NCP) changes. For NCP-specific static # routes, the '-p' option of the route(1M) command should be used. # # See netcfg(1M) for information about network configuration profiles. #
TIMEZONE: Asia/Kolkata
Checking disk-space for installation
Ok
Installing in /solaris10/zone10 ...
100% [===============================>]
Using a static exclusive-IP
./setupSysidcfgStaticIP[50]: file/sysidcfg.zone10: cannot create [No such file or directory]
cp: cannot access file/sysidcfg.zone10
chmod: WARNING: can't access file/sysidcfg.zone10
Attaching zone10
Booting zone10
zoneadm: zone 'zone10': must be installed before boot.
Waiting for boot to complete
booting...
booting...
booting...
root@localhost:# zoneadm list -cv
ID NAME STATUS PATH BRAND IP
0 global running / solaris shared
4 zone1 running /zone1 solaris excl
5 zone2 running /zone2 solaris excl
- zone10 configured /solaris10/zone10 solaris10 excl
root@localhost:#
I have installed the package pkg:/system/zones/brand/brand-solaris10 on the system.
# cat /etc/release
Oracle Solaris 11.1 SPARC
Copyright (c) 1983, 2012, Oracle and/or its affiliates. All rights reserved.
Assembled 19 September 2012

Sorry, please ignore my previous message about the wrong ip address (I assume you entered a valid ip address).
The problem you are facing is related to the defaultrouter, and I assume you are running Solaris 11 update 1.
If so, to workaround the problem for the time being you may edit the /etc/defaultrouter file in the global-zone and enter the ip address of the default router (you can remove this change when the template deployment is completed).
If you have no idea about the default router ip address, just rename /etc/defaultrouter (e.g. to /etc/defaultrouter.orig) so that it cannot be found by the template. Again, you can undo the change when deployment is complete.
Thierry

Whole root zone creation

Hi
I am building a whole root zone running Solaris U9 I noticed the   zone does not get installed with some updated packages from global zone like the openssl installed in global zone does not get installed in the non global zone and it also requires sshd user to be defined This results the ssh not starting in the new zone I want to know if this is the default behaviour of whole root zone installation process and how can we get the zone to be installed with all the currently running packages of the global zone Are there any post installation steps that need to be followed for the zone to boot up correctly. I have run the post post installation configuration steps as well
TIA

Thanks Walter
I was wondering since most of the packages installations at least require one interaction   where it asks for confirmation. how the other packages are getting installed. But mostly that might be the reason
There is also the SUNW_PKG_ALLZONES and other related variables   which control this but since here the packages which are already installed on global zome are getting installed on NG zone , how these parameters can be changed I could not found any of these parameters in any pkginfo file in the global zone

Zone creation

Hi,
My machine is x86 , OS is winxp
Iam using solaris10 running on VMware player. I have created a test zone on solaris10 using solaris9 container and down loaded solaris9-image.flar and saved on Apps folder. But iam unable to install solaris9-image.flar to test zone.
#zonecfg -z testzone info
zonepath: /zones/testzone
autoboot:true
pool:
inherit-pkg-dir:
dir: /lib
inherit-pkg-dir:
dir: /platform
inherit-pkg-dir:
dir: /sbin
inherit-pkg-dir:
dir: /usr
net:
address:
Physical:
#zoneadm -z testzone install -u -a /Apps/solaris9-image.flar
-u: illegal option --u
usage: install
install the configuration on to the system
As iam new in solaris can anyone help on regard this......Thanks in advance.
Arunii2

one more thing, I just noticed, that you are using X86 solaris. so have a look at this:
"A Solaris 9 branded zone (solaris9) is a complete runtime environment for Solaris 9 applications on *SPARC* machines running the Solaris 10 8/07 Operating System or later."
http://docs.sun.com/app/docs/doc/820-4490/gfjbk?a=view
so, you won't be able to test those things on x86 machine.
Laszlo

Zone creation(zonecfg) fails

I had created a few zones as root, and now when i try to create zones as a non-root user , it errors out
#/usr/sbin/zonecfg -z zone-appconfig
WARNING: you do not have write access to this zone's configuration file;
going into read-only mode.
zonecfg:zone-appconfig> create
zone-appconfig: cannot create in read-only mode.
#id
uid=150(user) gid=1(other)
Following are the zones that i had created as root
#/usr/sbin/zoneadm list -cv
ID NAME STATUS PATH
0 global running /
1 zone-small running /export/home/zone-small
- zone-whole configured /export/home/zone-whole

Ok..i could reach one more step by doing the following
.../etc/zones wasnt setup as world writable
# ls -l |grep zones
drwxr-xr-x 2 root sys 512 Jan 2 13:25 zones
# chmod g+w zones
# chmod o+w zones
# ls -l |grep zones
drwxrwxrwx 2 root sys 512 Jan 2 13:25 zones
zone-appconfig: No such zone configured
Use 'create' to begin configuring a new zone.
zonecfg:zone-appconfig> create
zonecfg:zone-appconfig> set zonepath=/home/abc
zonecfg:zone-appconfig> verify
zonecfg:zone-appconfig> exit
zone-appconfig: Problem updating index file
Configuration not saved; really quit (y/[n])? n
zonecfg:zone-appconfig> commit
zone-appconfig: Problem updating index file
zonecfg:zone-appconfig>
This is probably index file somewhere which isnt writable by my non-user account :(
Any help is appreciated...thx!!

Skipping sys-unconfig on Zone Creation

I guess I haven't created a new zone in awhile, just be using the 5 that I'd had for ages. I recently created a new Zone and fired it up but was supprised when I could log into it. When connecting to the console I was greeted with the familiar CURSES inferface for initial network setup. Are Zones now sys-unconfig'ing themselves during the install or ready phases? Is there a way to avoid having to do this?

I guess I haven't created a new zone in awhile, just
be using the 5 that I'd had for ages. I recently
created a new Zone and fired it up but was supprised
when I could log into it. When connecting to the
console I was greeted with the familiar CURSES
inferface for initial network setup. Are Zones now
sys-unconfig'ing themselves during the install or
ready phases? Is there a way to avoid having to do
this?See sysidcfg(4) -- you can place such a file in the zone (at /path/to/zone/root/etc/sysidcfg) before you boot it for the first time, and the sysid tools will pick up the settings from it.
- jonathan

Independent Zone Paths for Failover Services

We wish to use Sun Cluster 3.2 to operate failover zones across 3 or more servers. It would seem that for failover services, zone paths must exist on shared storage and all associated failover zones must use that exact zone filesystem. However for load-balancing, it appears that multiple independent zonepaths are allowed.
My question is can failover zones be deployed WITHOUT using a single zone path filesystem. It seems to me that it complicates patching and maintenance to the zone filesystem if it is always in use. Not to mention, a corruption in that zone path's files would be forced on the entire set of zones using that path.
I'm new to Sun Cluster and am skittish about deploying a large fiber channel infrastructure if it is not really necessary. Thanks for any corrections / advice on the above.

Now I understand what you want to achieve. Unfortunately wording is not yet precise with all these new technologies.
What you need is not the HA Container agent, and what you want is not failover zones. But you want to have local zones, and let HA services failover between local zones on different nodes. Nothing easier than that. Just use the "normal" procedures to configure an agent. Now, instead of using the nodenames for the nodelist of the resource group, give it a list of the zone names that you want this service to run in. Example could be:
clrg create -p nodelist=host-A:zone-FOO,host-B:zone-FOO,host-C:zoneFOO. Now you can switch the HA service between the zone-FOO(s) on the 3 nodes without switching the zones themselves. And you can patch them without taking the service offline. Great.
If you want to make the data that your HA service uses highly available, you still would need some kind of shared filesystem. How could the HA service, switched onto another node:zone would get access to it, if not via a filesystem that was also switched over?
There is still value to the HA COntainer agent: if, e.g. you want to use the blackbox approach and do not care much about the potential outages of the zones during patch cycles. And there is more.
Hope that helps
Hartmut

Physical Connectivity from Fabric Interconnect to MDS and Failover please suggest

Dear Team
We have 2 FI and 2 MDS 1 SAN
Currenlty the connectivity is
2 direct physical connections from FI-A to MDS1
2 direct physical Connections from FI-B to MDS2
From MDS 1 connection to SAN Primary controller
From MDS 1 connection to SAN Secondary controller
From MDS 2 connection to SAN primary controller
From MDS 2 connection to SAN Secondary controller
Hope the above connectivty is fine?
we had looked at
http://www.cisco.com/en/US/prod/collateral/ps4159/ps6409/ps5990/white_paper_c11_586100.html (as team always preferred to go via cisco standards)
In FI we have to cluster the FI and Primary and Subordinate,likewise here for MDS to we need to? or will it be done via FIs ?
If not these 2 MDs they work independent but sharing the configs that happens among them?
Now what ever changes (Zone Creation)we make on MDS1 hope that will get relpicated to MDS2 via FIs?
If MDS1 Fails will all the configuration be available on MDS2 and still the infra will be smooth running ? and vice versa?
Any additional steps to be done to achieve it?
Which is the best way to achieve it please suggest
Thanks and Regards
Jose

Hi Jose.
Physically your connectivity is good.
On the UCS side, the 'Primary' and 'Subordinate' role refer only to the Management of the system and which device is actually running UCS Manager.
Each MDS device will have separate configuration (zoning). It's different between the 2 devices.
The blade itself will have a connection out each side, or 'SAN Fabric'.
e.g.
           +----+
       +---+    +-----+
       |   |    |     |
       |   |    |     |
       |   |SAN |     |
       | +->----<--+ |
       | |         | |
       | |         | |
   +---+-+-+    +--+--+--+
   |       |    |        |
   |MDS-A |    |MDS-B   |
   +--^-^--+    +---^-^--+
      | |           | |
   +--+-+--+    +---+-+--+
   |       |    |        |
   |UCS-A |    | UCS-B |
   +--+----+    +-----+--+
      |               |
      |               |
VSAN100    +-----+    | VSAN200
      |    |     |    |
      +---->     <----+
           |Blade|
           +-----+
The blade will have a HBA on FabricA (VSAN100) and FabricB (VSAN200)
Each HBA will have a different WWPN, and on the SAN array, each controller will have a WWPN
So on MDS-A, the zoning will be:
Blade WWPN A
Storage WWPN Primary A
Storage WWPN Secondary A
MDS-B, the zoning will be:
Blade WWPN B
Storage WWPN Primary B
Storage WWPN Secondary B
So the configuration is *not* synced between the two MDS devices, but they each have visibility to the blades vHBA device. At the blade level, the Multipathing software on the Operating System will handle any failover.
On the UCS, we would generally use a Port Channel up to the MDS.

Failover Zone Creation

Similar Messages

Maybe you are looking for