Description of problem: 1. Get a quorate cluster up 2. attempt a ccs_connect and it will return a vaild cd 3. update the cluster.conf file on another node and send a HUP 4. wait 5. a ccs_connect attempt will fail 6. a second ccs_conncet attempt will return a vaild cd [root@morph-05 root]# cat /proc/cluster/nodes Node Votes Exp Sts Name 1 1 5 M morph-05 2 1 5 M morph-02 3 1 5 M morph-04 4 1 5 M morph-03 5 1 5 M morph-01 [root@morph-05 root]# ccs_test connect Connect successful. Connection descriptor = 0 [root@morph-05 root]# ccs_test connect Connect successful. Connection descriptor = 1 ### Updated file on another node ### the new file appears: [root@morph-05 root]# head /etc/cluster/cluster.conf <?xml version="1.0"?> <cluster name="morph-cluster" config_version="5"> <cman> </cman> <dlm> </dlm> <nodes> [root@morph-05 root]# ccs_test connect ccs_connect failed: Operation not permitted [root@morph-05 root]# ccs_test connect Connect successful. Connection descriptor = 2 SYSLOG: Oct 4 16:36:24 morph-05 ccsd[2138]: cluster.conf on-disk version is <= to in-memory version. Oct 4 16:36:24 morph-05 ccsd[2138]: On-disk version : 5 Oct 4 16:36:24 morph-05 ccsd[2138]: In-memory version : 5 Oct 4 16:36:24 morph-05 ccsd[2138]: Failed to update config file, required by cluster. Oct 4 16:36:24 morph-05 ccsd[2138]: Error while processing connect: Operation not permitted How reproducible: Always
Updates with the proper version and component name.
Updates with the proper version and component name. Again, just love out tools.
- fix bug 143165, 134604, and 133254 - update related issues These all seem to be related to the same issue, that is, remote nodes were erroneously processing an update as though they were the originator - taking on some tasks that didn't belong to them. This was causing connect failures, version rollbacks, etc.
fix verified.