High Availability

[Darren Williams]

Does anyone have any working scenarios they can give examples of please?

With two datacenters I'm currently thinking along the lines of:

A redis instance in AWS

DC1

Opensips1, Session Manager1, Rals1

DC2

Opensips2, Session Manager2, Rals2

I think if I make everything go to Rals1 first that with Rals2 as failover that would be OK?

My concern is latency between the Rals instances and also redis.

It would be good to hear how others are handling this.

Thanks

Daz

[DanB]

Hey Daz,

for the moment, the best way is to have master/slaves configured in
"_conns" and let CGRateS deal with connecting/reconnecting
automatically between two locations.
Regarding latency, it all depends on how far distance is between your
datacenters but if voice goes well, I would not worry about RPC calls.

DanB

> --
> You received this message because you are subscribed to the Google
> Groups "CGRateS" group.
> To unsubscribe from this group and stop receiving emails from it,
> send an email to cgrates+u...@googlegroups.com.
> To post to this group, send email to cgr...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/
> msgid/cgrates/bd1f9e79-9c26-4860-a317-219fbdbc6137%40googlegroups.com
> .
> For more options, visit https://groups.google.com/d/optout.