Hi, Denys
Talking about HA ,it's complex , and have lots of detail.
The "FS HA Solution" is very simple . if you think deeper ,you will find more question,such as:
A. if we have 5 servers in the group , what is the rule for transfer the calls ?
B. Once the broken server have 1000 concurrent call , if there's new "performance issue" when we transfer so many calls to a server.
If it really need an answer about your question -- "if it is possible to move calls back". I think it's unnecessary. Every time when we transfer the calls from one server to another, there's a little "audio dropped" , maybe , not good for user experience.
So, forgot your question , the "HA Solution" is just a demo . It need more work to do for product situation.
Raymond
ÔÚ 2017-06-01 06:28:42£¬"Denys Pozniak" <
denys.pozniak@crazycall.com> дµÀ£º
Hello!
I built FS HA solution based on keepalived and mysql master-master.
It works ok generally, but as I understand FS after restarting cleaning own database.
So when node1 fails calls jump to node2, after script restarts node1 it is not possible to move calls back.
Tried options in switch.conf.xml, but no luck:
<param name="auto-create-schemas" value="false"/>
<param name="auto-clear-sql" value="false"/>
Is there is a way to solve this?
BR,
Denys