thanks

excessive paging resolution is simple, add memory or reduce load.

check if you can reach other primary node from secondary, check endpoint config on all nodes. , check firewall/ports

modify the code as per your requirement

declare @cmd varchar(max)
declare @cmd_suspend varchar(max)
declare @cmd_resume varchar(max)
declare c1 cursor read_only for
select 'ALTER DATABASE ['+DB_name(database_id)+']...

why  only page compression, where is compression best practice ?

must add route between subnets and/or ip addresses , talk to you network admin, he/she will understand.

it should work,  I see no reason for it not to work.

based on scenario you described  you need to assign both ips to the clustername , you can have...

if it is virtual , then may be add virtual nic... with 10.10.25.x 
however same steps are required as above just on new nics.

add additional  IP to each node with 10.10.25.x.
add additional IP to cluster with 10.10.25.x.
then you can add 10.10.25.x. to listener. 

keep your existing IP addresses as well.

gud luck

run proc-explorer  (https://docs.microsoft.com/en-us/sysinternals/downloads/process-utilities ) or task manager (maybe)  to monitor what application is doing backup..  (you may look at timestamp of the backups to find when should you run...

you can achieve this if you design system  using one table per database, it is crazy thou, good luck 🙂

you can also bcp out each table.

well this was a copy of a prod database, that i used to recover some data, the database was restored with stopat timestamp and recoverd

yes and it did work, however I am not able to find what caused it

this is quite old, but you can try that , hope it helps , if you decide to go ahead from 2012 to 2017 be ready, you might be declined...

Perfect example of disaster, inability of DBA not to implement correct backup strategy and/or not having control over database backups, This is what happens if you give control of backup...