Wrong POD-CIDR when recovery from lost contiv-etcd

[LipingMao]

Sometimes(not every time), if I delete contiv-etcd, the POD-CIDR will overlap. Here is an example: ams-dev-limao-work002 and ams-dev-limao-work004 get the same POD CIDR with 10.1.1.0/24.

[root@ams-dev-limao-ctrl001 ~]# contiv-netctl ipam

time="2019-03-29 12:58:26.41316" level=info msg="Connected to Etcd (took 1.28459ms)" endpoints="[127.0.0.1:32379]" loc="etcd/bytes_broker_impl.go(57)" logger=defaultLogger
ID  NODE-NAME              VPP-IP       BVI-IP        POD-CIDR     VPP-2-HOST-CIDR  POD-CLUSTER-CIDR
2   ams-dev-limao-ctrl001  10.253.3.45  192.168.30.2  10.1.2.0/24  172.30.2.0/24    10.1.0.0/16
7   ams-dev-limao-ctrl002  10.253.3.50  192.168.30.7  10.1.7.0/24  172.30.7.0/24    10.1.0.0/16
3   ams-dev-limao-ctrl003  10.253.3.44  192.168.30.3  10.1.3.0/24  172.30.3.0/24    10.1.0.0/16
5   ams-dev-limao-work001  10.253.3.32  192.168.30.5  10.1.5.0/24  172.30.5.0/24    10.1.0.0/16
4   ams-dev-limao-work002  10.253.3.46  192.168.30.4  10.1.1.0/24  172.30.1.0/24    10.1.0.0/16
6   ams-dev-limao-work003  10.253.3.27  192.168.30.6  10.1.6.0/24  172.30.6.0/24    10.1.0.0/16
1   ams-dev-limao-work004  10.253.3.37  192.168.30.1  10.1.1.0/24  172.30.1.0/24    10.1.0.0/16

[rastislavs]

Is this issue still present in the latest Contiv-VPP version? I wasn't able to reproduce it.