I fixed the mcast addr. It worked (Vadym Chepkov also responded with the mcast addr issue). Thanks!
Now, when I run the tests, I get the following: [r...@localhost test]# ./testcpg2 All of the nodeids should match on a single node configuration for the test to pass.local_get: d9dd17ac confchg nodeid d9dd17ac self delivered nodeid: d9dd17ac [r...@localhost test]# ./testcpg Local node id is d9dd17ac membership list node id -639821908 pid 977 Type EXIT to finish ConfchgCallback: group 'GROUP' joined node/pid -639821908/977 reason: 1 nodes in group now 1 node/pid -639821908/977 From the above output, I'm inferring that the test is successful. However, I'm not sure how I should interpret the pid in the node/pid output of testcpg. The "pid" of corosync is 945 and not 977: [r...@localhost test]# ps -ael | grep corosync 1 S 0 945 1 0 -40 - - 35751 209558 ? 00:00:00 corosync After this, I went ahead and downloaded openais-1.1.3.tar.gz. I then did the following successfully: # configure # make install I tried running the following test program: [r...@localhost test]# ./testclm Could not initialize Cluster Membership API instance error 12 Running other test programs gives similar errors. What should I be doing to get these tests to pass? Is there any other setup required. Thanks, - Kishore -----Original Message----- From: Steven Dake [mailto:sd...@redhat.com] Sent: Saturday, May 22, 2010 11:07 AM To: Sampathkumar, Kishore Cc: openais@lists.linux-foundation.org Subject: Re: [Openais] corosync is having problems on a single node On Sat, 2010-05-22 at 21:57 +0530, Sampathkumar, Kishore wrote: > I’m trying to setup corosync on a single node. > > > > I downloaded corosync-1.2.3.tar.gz, and after extracting it, issued > the following as root on a single node: > > # configure > > # make install > > > > The above succeeded. > > > > I then did the following: > > # cd /etc/corosync > > # cp corosync.conf.example corosync.conf > > > > I changed the following lines thus: > > bindnetaddr: 172.23.220.0 > > mcastaddr: 172.23.223.255 > This multicast address is invalid. A multicast address starts at 224.x.x.x and runs through 239.x.x.x. Try resolving that first. Regards -steve > Other than the above, I have not changed anything in corosync.conf > > > > [ The output from ifconfig on this system is shown below: > > [r...@localhost corosync]# ifconfig > > eth0 Link encap:Ethernet HWaddr 00:18:8B:39:6C:FB > > inet addr:172.23.221.217 Bcast:172.23.223.255 > Mask:255.255.252.0 > > inet6 addr: fe80::218:8bff:fe39:6cfb/64 Scope:Link > > UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 > > RX packets:36179218 errors:0 dropped:0 overruns:0 frame:0 > > TX packets:51451 errors:0 dropped:0 overruns:0 carrier:0 > > collisions:0 txqueuelen:1000 > > RX bytes:2904591110 (2.7 GiB) TX bytes:4540059 (4.3 MiB) > > Interrupt:169 Memory:f8000000-f8012800 > > > > lo . . . ] > > > > After the above, I ran the following as root successfully: > > # corosync-keygen > > > > I started the corosync executive as follows: > > # corosync > > > > The above started fine. I re-verified using ps: > > [r...@localhost corosync]# ps -ael | grep corosync > > 5 S 0 639 1 0 -40 - - 35494 209558 ? 00:00:00 > corosync > > > > Now, I tried running the supplied test program with corosync (under > test directory): > > [r...@localhost test]# ./testcpg > > Local node id is d9dd17ac > > Could not join process group, error 6 > > [r...@localhost test]# ./testcpg2 > > All of the nodeids should match on a single node configuration > > for the test to pass.local_get: d9dd17ac > > testcpg2: testcpg2.c:81: main: Assertion `CS_OK==cpg_join(handle, > &group)' failed. > > Aborted > > > > What should I be doing to get this to run successfully? > > > > Thanks, > > - Kishore > > > _______________________________________________ > Openais mailing list > Openais@lists.linux-foundation.org > https://lists.linux-foundation.org/mailman/listinfo/openais _______________________________________________ Openais mailing list Openais@lists.linux-foundation.org https://lists.linux-foundation.org/mailman/listinfo/openais
