If you restore from a nim server with an LPP_source assigned any missing devices will be installed. 1. I highly recommend you make a mksysb of the system as it is, and set that somewhere safe. 2. Using alt_disk_install, e.g., do an update_all within the TL level you are running, or update to the latest TL. Updating a TL should ensure that all device drivers (also the currently unused virtual drives) are installed. If something is missed though, nim procedure will install what is missing. 3. Boot to new image, do some basic tests to be sure it is working as expected, and make an mksysb. This is the image you will clone.
4. When you restore from the nim server do not restore old devices - as all you old ethernet and i/o adapters will not be there. Depending on your network settings in the old situation you may have some additional work getting all the interfaces as expected. 5. If the disks are coming from a LUN now (but direct rather than as NPIV or VSCSI) they should work as expected. However, any tuning you may have done to the adapters or hdisks (number of commands, no_reserve, queue_depth, etc) will need to be changed again. Note: some of these changes will be on the VIOS (physical adapter settings) and some in the client.
It is fairly straightforward, but the first time will seem like a lot of activity. You may want to test it just to have a practice run first. p.s. You can skip the update_all first, but as you are getting some planned downtime to do this it is an excellent time to get an update done as well!
I do have a scenerio where we have to clone one existing P570 ( POWER5) with dedicated Lpars ( no VIO)to an about to purchase P520 ( POWER 6 ) with virtulaized Lpars.
we have decided to go for mksysb restoration technique , but do have following questions:
1, Are there any be issue expected while restoring mksysb of a POWER 5 server to POWER 6 server ( AIX 5.3 will be the OS for both source/target servers)
2. Will there any be issue expected while restoring mksysb of a dedicated resources Lpar to a virtualized resources Lpar. offcourse we will be setting VIO with new p520 separately?
The problem is solved. I used userattrmappath:/etc/security/ldap/2307user.map in ldap.cfg and the user attribute name mapping table not match with LDAP server user attribut.
speed workaround: I modified /etc/security/ldap/2307user.map pgrp SEC_CHAR gid s groups SEC_LIST grouplist m
I modified /etc/security/ldap/2307group.map keyobjectclass SEC_CHAR aixaccessgroup s groupname SEC_CHAR groupname m id SEC_INT groupname s users SEC_LIST member m
I'll try to use 2307aixgroup.map and 2307aixuser.map with some modification.
Hi. It has been a long while since I have worked with LDAP on AIX, but I just want to verify whether you removed the LDAP client and then reran the mksecldap command with the new scheme as argument. Otherwise I suspect the default scheme being used is still the AIX scheme.
I'll try and get an answer for someone I know who is more familiar with LDAP ins and outs. But purely from an installation point of view I would want to verify the exact steps and commands you used to get to the current state.
I install Linux Debian 5.04 and slapd package (LDAP server) and running fine. all user can login on linux. On AIX 5300-11-02, i install ldap client fileset from AIX DVD, I can not see the pgrp and groups from user002 with lsuser command and login failed, i use AIXaccessGroup objectclass. However with user1, i use posixgroup objectclass and that work's fine. I add rfc2307aix and aixaccessgroup to /etc/security/ldap/ldap.cfg ... nothing.
Below my little TEST.ldif added to LDAP server and AIX lsuser commands result. user001 and tech group : OK user002 and grp1 + grp2 groups : BAD
TEST.ldif ------------------------------------ dn: ou=people,dc=domain,dc=local objectClass: top objectClass: organizationalUnit ou: people
dn: ou=group,dc=domain,dc=local objectClass: top objectClass: organizationalUnit ou: group
dn: cn=tech,ou=group,dc=domain,dc=local objectClass: posixGroup objectClass: top cn: tech gidNumber: 10000
Thanks Michael, I have read the redbooks, basically we need to define sites first with the nodes in the list and then create a cluster, we will then add sites into the cluster, a bit complex but not hard
if someone has any PowerHA failover scripts then that would be help, I can use them as a template, basically due to apps workload we need to create failover script so for instance if 2 or more than 2 nodes are down in the 6 nodes cluster then failover to the other site instead of 3rd node in the same cluster etc.
The new name is something like PowerHA SystemMirror (version 6.1 and beyond for now). I am not as active with HACMP/PowerHA as I once was.
Basically, it should be fairly simple with XD, i.e. all in smit screens - as XD knows how to talk with storage systems. If I recall correctly XD negotiates the visibility of the disks. On the side/site that is active - the disks are visible, on the inactive side they are invisible, or otherwise inaccessable. On a node failure, or directed move of a resource group XD/SystemMirror cause the visibility status to change.
Maybe visibility is the wrong term - but the "smarts" of CD is that it negotiates with the storage system where the LUNs are visible/active.
And today I just heard that there are two or three new courses from IBM training for HACMP/PowerHA/SystemMirror. The course codes are AN41-AN46 (some of these course are already a year old, others are less than a few weeks.)
When I have time I'll look a bit deeper - so maybe someone who has more recent experience will give an answer before that.
