04-26-2013
Little time - so I shall be quick.
What PowerHA/HACMP (now System Mirror) monitors - traditionally - is the topology - for failures - and based on what has happened, it makes modifications.
1) Disks are a resource, not topology
2) diskhb network is a NON-IP network, so application service is not dependent on that.
3) HACMP does not respond - out of the box - to resource failures. You need to setup an application and/or event notification to respond to that.
Suppose, rather than both fc - you had said both/all ethernet (i.e., IP connectivity) - the result would be the same - HACMP does not move the resource group. This is because HACMP can see both nodes are operational - but NO nodes (there are only two!) can communicate with other. When no nodes can communicate via IP network HACMP concludes a "Global" network failure, and since no nodes can communicate - it does not matter where the application/service is at - no clients could communicate - so nothing is moved.
The short of it - you need to configure an application monitor that "decides", whether to move or not. A simple way to force a move (one that would work "out of the box" is a so-called panic halt (halt -q). Not really recommended of course, but it does, in an educational/test phase help demonstrate what events HACMP responds to, and what events it does not respond to.
Hope this helps.
10 More Discussions You Might Find Interesting
1. AIX
(0 Replies)
Discussion started by: mhenryj
0 Replies
2. AIX
Hello
I am a noobie to HACMP
I have 2 55A servers and a 7031 disk subsystem
For HACMP to work, do I need to have the hdisks on both servers to match the same drives??
The hdisks on each box are currently different:
hdisk0 and hdisk1 are the internal disks and hdisk2-5 are on the D24 on... (1 Reply)
Discussion started by: mhenryj
1 Replies
3. HP-UX
Hi,
Can anybody advice whether it is possible to configure HACMP in HP-UX Server.
To my knowledge HACMP is IBM Solution.
Thanks && Regards,
N. Poorna Chandra Rao. (2 Replies)
Discussion started by: npcrao
2 Replies
4. AIX
Hi,
I needed space on a FS, and when I've added the space on the filesystem, I did it trough the regular smitty fs inteface and not with smitty cl_lvm.
Can someone help me to repair the situat before a faileover happen ?
Thanks for your help,:mad: (13 Replies)
Discussion started by: azzed27
13 Replies
5. AIX
Hi,
Can we use network for heartbeat, I mean can we use different network card for heartbeat. (6 Replies)
Discussion started by: vjm
6 Replies
6. AIX
hi
can anyone explain the concepts of HACMP and configuration (step by step) (2 Replies)
Discussion started by: udtyuvaraj
2 Replies
7. AIX
hi,
when I do a failover, hacmp always starts db2 but recently it fails to start db2..noticed the issue is db2nodes.cfg is not modified by hacmp and is still showing primary node..manually changed the node name to secondary after which db2 started immediately..unable to figure out why hacmp is... (4 Replies)
Discussion started by: gkr747
4 Replies
8. AIX
Does anyone has idea about, what is the ibm standard HACMP trip interval?
We have 20 second.
lssrc -ls topsvcs
Subsystem Group PID Status
topsvcs topsvcs 1843200 active
Network Name Indx Defd Mbrs St Adapter ID Group ID
HB Interval =... (7 Replies)
Discussion started by: allwin
7 Replies
9. AIX
Hi.
We have a two node HA cluster.
We got a request to change one of the VG name?
Is there an option to do this online ? If it requires downtime can someone please explain me the steps for doing it ?
Let me know if you need any outputs from the servers (1 Reply)
Discussion started by: newtoaixos
1 Replies
10. AIX
hello
ive 2 nodes aix 6100-04-10-1119 with hacmp 6.1.0.0
my qustion is :
can i add default gateway to one member node with hacmp a live or to do take over to realse one of the members ?
thanks
When it is urgent, then open the thread up in the appropriate subforum next time, thanks. (3 Replies)
Discussion started by: ariec
3 Replies
LEARN ABOUT CENTOS
migratepages
MIGRATEPAGES(8) Linux Administrator's Manual MIGRATEPAGES(8)
NAME
migratepages - Migrate the physical location a processes pages
SYNOPSIS
migratepages pid from-nodes to-nodes
DESCRIPTION
migratepages moves the physical location of a processes pages without any changes of the virtual address space of the process. Moving the
pages allows one to change the distances of a process to its memory. Performance may be optimized by moving a processes pages to the node
where it is executing.
If multiple nodes are specified for from-nodes or to-nodes then an attempt is made to preserve the relative location of each page in each
nodeset.
For example if we move from nodes 2-5 to 7,9,12-13 then the preferred mode of operation is to move pages from 2->7, 3->9, 4->12 and 5->13.
However, this is only posssible if enough memory is available.
Valid node specifiers
all All nodes
number Node number
number1{,number2} Node number1 and Node number2
number1-number2 Nodes from number1 to number2
! nodes Invert selection of the following specification.
NOTES
Requires an NUMA policy aware kernel with support for page migration (linux 2.6.16 and later).
migratepages will only move pages that are not shared with other processes if called by a user without administrative priviledges (but with
the right to modify the process).
migratepages will move all pages if invoked from root (or a user with administrative priviledges).
FILES
/proc/<pid>/numa_maps for information about the NUMA memory use of a process.
COPYRIGHT
Copyright 2005-2006 Christoph Lameter, Silicon Graphics, Inc. migratepages is under the GNU General Public License, v.2
SEE ALSO
numactl(8) , set_mempolicy(2) , get_mempolicy(2) , mbind(2) , sched_setaffinity(2) , sched_getaffinity(2) , proc(5) , ftok(3) , shmat(2) ,
taskset(1)
SGI
Jan 2005 MIGRATEPAGES(8)