The patch, even with the keepalive settings, was not working for me on-site at customer so I added crm configuration to restart maas-regiond whenever pgsql gets promoted to master.
primitive res_maas-regiond systemd:maas-regiond \
op start interval=0 timeout=120s \
op monitor interval=10s timeout=120s \
op stop interval=0 timeout=120s
clone cl_maas-regiond res_maas-regiond
order ord_maas_restart Mandatory: \
ms_pgsql:promote \
cl_maas-regiond:start \
symmetrical=false
This resource is cloned to all nodes and then tied to the promotion of any ms_pgsql node. If a node in ms_pgsql is promoted, it will then restart maas-regiond (via systemd) on all nodes.
I feel this is solving the problem with a hammer but it works.
The patch, even with the keepalive settings, was not working for me on-site at customer so I added crm configuration to restart maas-regiond whenever pgsql gets promoted to master.
primitive res_maas-regiond systemd: maas-regiond \
op start interval=0 timeout=120s \
op monitor interval=10s timeout=120s \
op stop interval=0 timeout=120s
clone cl_maas-regiond res_maas-regiond
order ord_maas_restart Mandatory: \ pgsql:promote \ maas-regiond: start \ =false
ms_
cl_
symmetrical
This resource is cloned to all nodes and then tied to the promotion of any ms_pgsql node. If a node in ms_pgsql is promoted, it will then restart maas-regiond (via systemd) on all nodes.
I feel this is solving the problem with a hammer but it works.