Operating System: All
Platform: All

> following error:

Here [1] is the raw data from topology-netconf. You can see that there is one device with lower port number, stuck in "connecting" status:

{"node-id":"netconf-test-device-17831-5","netconf-node-topology:available-capabilities":{},"netconf-node-topology:host":"10.29.13.99","netconf-node-topology:unavailable-capabilities":{},"netconf-node-topology:connection-status":"connecting","netconf-node-topology:port":17831}

Configurer output [2]:
(delete:200)x102, (put:201)x123
means the expected number of devices to be seen is 21, presumably implying netconf-test-device-17831-5 has been deleted from config datastore, but not from topology.

There is also log from testtool [3], but it is debug level, so not easy to read.

> we should also check the correctness of the suite, because now I am not sure how it exactly works.

Most of the work is done by the configurer script [4]. Previously the suite was quite broken, but I think it works correctly now.

[1] https://logs.opendaylight.org/releng/jenkins092/netconf-csit-3node-cluster-stress-all-nitrogen/46/log.html.gz#s1-s2-t7-k2-k1-k1-k1-k4-k1
[2] https://logs.opendaylight.org/releng/jenkins092/netconf-csit-3node-cluster-stress-all-nitrogen/46/configurer--netconf-cluster-stress-txt-Topology-Owner-Ha.log.gz
[3] https://logs.opendaylight.org/releng/jenkins092/netconf-csit-3node-cluster-stress-all-nitrogen/46/testtool--netconf-cluster-stress-txt-Topology-Owner-Ha.log.gz
[4] https://github.com/opendaylight/integration-test/blob/master/tools/netconf_tools/configurer.py