FilipBB

Hello, wondering if there is any progress on this issue? I cannot migrate from 1.0 > 2.0 because this error comes up during import.

Sure, thanks for the feedback Manolo. It's quite hacky because I don't have formal programming experience, so I don't know best practices. Let me know when you're done looking it...

I ran the command that the ansible task runs on the replica, on the replica, namely: ``` /usr/bin/pg_ctlcluster 16 main start -o '-c hot_standby=off' ``` The log I posted is...

Maybe it's because I am using a custom data dir? ``` root@pgnode1:/home/ubuntu# /usr/bin/pg_ctlcluster 16 main start -o '-c hot_standby=off' Notice: extra pg_ctl/postgres options given, bypassing systemctl for start operation The...

Ok, will do, thanks for looking into it so quickly.

It's strange, it gets stuck on starting the database on the master node. It seems to be stuck in a loop of restoring and trying to start up but I...

The intended master seems to be starting up as a standby according to the patroni logs, which correlates with what I can find on google for the "postgres@postgres FATAL: the...

Yeah, it seems like this node is not designated as the master for some reason, but I have not changed variables or anything else from 1.9.0.

FYI this is the inventory file: ``` [etcd_cluster] 10.60.1.154 ansible_ssh_host= ansible_ssh_port=2231 10.60.1.155 ansible_ssh_host= ansible_ssh_port=2232 10.60.1.156 ansible_ssh_host= ansible_ssh_port=2233 [master] 10.60.1.154 ansible_ssh_host= ansible_ssh_port=2231 hostname=pgnode0 postgresql_exists=false [replica] 10.60.1.155 ansible_ssh_host= ansible_ssh_port=2232 hostname=pgnode1 postgresql_exists=false 10.60.1.156...

So it seems like etcd elected pgnode1 (10.60.1.155) to be master, instead of what I set in ansible to be the master (pgnode0 10.60.1.154): ``` root@pgnode0:/home/ubuntu# etcdctl endpoint status --cluster...