Snapshots work fine. I can make a snapshot -> convert it to template
and start the VM on a new node from that template. When I needed to
move one VM for balance purposes. But I want to fix the migration
process. I have attached the error log to this email. Maybe I'm
looking at the wrong place were
Hi Darius,
>From the logs, I could observe that image volume is already in the creating
state and trying to use the same for copying the volume between pools. So,
state transition failed. Could you please provide the complete log for the
usecase to root cause the issue. Also, include volumes and s
Hi Rohit,
thx a lot for sharing that - here, if I understand correctly, you relly on
the static IP (range) for the systemVM (4.11) - right - and then use LB on
top of that...
But any viable solution for pre-4.11 releases, where CPVM can get any of
the public IPs - LB is possible but even worse th
Hi Andrija,
We use your second option – “create a LB (single public IP and thus single DNS
A entry), BUT do loadbalancing on top of 1000 backend public IPs”.
This depends on which load balancer you use and what functions this has
available – but in our case, we simple set up the full public rang
Hi Dag,
sorry for being a noob - but if you have 1000 (even 200 only) Public IPs,
do you have some special option (on your special LB) to add :backend/proxy
target in a form of a CIDR range, or you need, like in HaProxy to make 1000
lines with 1000 possible server backedns... ?
Sorry for boring
Hi Andrija,
This is actually using a Netscaler VPX which has the health check function
built in and allows for a simple IP range specification.
Looking at the HAproxy docs it doesn’t give the impression it can do the same
so I suspect you may have to just hand crank the config line by line – n
