06-09-2013
It feels like the synchronization was performed - "forced". As Bakunin mentions, it is most likely a communication problem.
Basic test - start all nodes but do not start a resource groups. I suspect you will not be able to start all nodes.
10 More Discussions You Might Find Interesting
1. Shell Programming and Scripting
Hello,
Under ksh I have to run a script on one of the nodes of a Solaris 8 cluster which at some time must execute a command on the alternate node:
# rsh <name> "command"
I have to implement this script on all the clusters of my company (a lot of...).
Fortunately, the names of the two nodes... (11 Replies)
Discussion started by: heartwork
11 Replies
2. HP-UX
Need help guys!
when running cmrunnode batch i'm getting this error
cmrunnode : Waiting for cluster to... (1 Reply)
Discussion started by: Tris
1 Replies
3. HP-UX
Hi,
Please advise me whereas I have two node cluster server configured with MC/SG. Application and DB are running on Node 1, while Node 2 is standby.
All the volume group devices are part of cluster environment. There is only one package running at node 1.
Node 2 is having the problem to... (1 Reply)
Discussion started by: rauphelhunter
1 Replies
4. High Performance Computing
hi,
i am trying to setup a 2 node cluster environment. following is what i have;
1. 2 x sun ultra60 - 450MHz procs, 1GB RAM, 9GB HDD, solaris 10
2. 2 x HBA cards
3. 2 x Connection leads to connect ultra60 with D1000
4. 1 x D1000 storage box.
5. 3 x 9GB HDD + 2 x 36GB HDD
first of all,... (1 Reply)
Discussion started by: solman17
1 Replies
5. Solaris
Hi Gurus,
For learning purpose, I have installed a single node cluster 3.2 on Solaris 10 for practice. Now I am welling to create two non-global zone and create them as a fail over. Will appreciate your help and assistance.
Thanks (3 Replies)
Discussion started by: newadmin
3 Replies
6. Solaris
I now the logical name and Virtual IP of the cluster.
How can I find the active sun cluster node without having root access? (3 Replies)
Discussion started by: sreeniatbp
3 Replies
7. HP-UX
Hello,
Is there any way to identify the active node in a HP-UX cluster without root privileges? (3 Replies)
Discussion started by: psimoes79
3 Replies
8. Solaris
Hi Gurus,
I am very new to clustering and for test i have created a single node cluster, now i want to remove the system from cluster. Did some googling however as a newbee in cluster unable to co related the info.
Please help
Thanks (1 Reply)
Discussion started by: kumarmani
1 Replies
9. Solaris
How can we add a shared zfs dataset between 2 zones on a same host. I have sun cluster 3.2 installed in a server which has 2 zones. I want to share a zfs data set between these 2 zones how can we do that ? (7 Replies)
Discussion started by: fugitive
7 Replies
10. Solaris
Hi,
Is it possible to have a Solaris cluster of 2 nodes at SITE-A using SVM and creating metaset using say 2 LUNs (on SAN). Then replicating these 2 LUNs to remote site SITE-B via storage based replication and then using these LUNs by importing them as a metaset on a server at SITE-B which is... (0 Replies)
Discussion started by: dn2011
0 Replies
MCKEY(1) librdmacm MCKEY(1)
NAME
mckey - RDMA CM multicast setup and simple data transfer test.
SYNOPSIS
mckey -m multicast_address [-s] [-b bind_address] [-c connections]
[-C message_count] [-S message_size] [-p port_space]
mckey -m multicast_address -s [-b bind_address] [-c connections]
[-C message_count] [-S message_size] [-p port_space]
mckey -M unmapped_multicast_address -b bind_address [-s] [-c connections]
[-C message_count] [-S message_size] [-p port_space]
DESCRIPTION
Establishes a set of RDMA multicast communication paths between nodes using the librdmacm, optionally transfers datagrams to receiving
nodes, then tears down the communication.
OPTIONS
-m multicast_address
IP multicast address to join.
-M unmapped_multicast_address
RDMA transport specific multicast address to join.
-s Send datagrams to the multicast group.
-b bind_address
The local network address to bind to.
-c connections
The number of QPs to join the multicast group. (default 1)
-C message_count
The number of messages to transfer over each connection. (default 10)
-S message_size
The size of each message transferred, in bytes. This value must be smaller than the MTU of the underlying RDMA transport, or an
error will occur. (default 100)
-p port_space
The port space of the datagram communication. May be either the RDMA UDP (0x0111) or IPoIB (0x0002) port space. (default
RDMA_PS_UDP)
NOTES
Basic usage is to start mckey -m multicast_address on a server system, then run mckey -m multicast_address -s on a client system.
Unique Infiniband SA assigned multicast GIDs can be retrived by invoking mckey with a zero MGID or IP address. (Example, -M 0 or -m
0.0.0.0). The assigned address will be displayed to allow mckey clients to join the created group.
Because this test maps RDMA resources to userspace, users must ensure that they have available system resources and permissions. See the
libibverbs README file for additional details.
SEE ALSO
rdma_cm(7), ucmatose(1), udaddy(1), rping(1)
librdmacm 2007-05-15 MCKEY(1)