> That would be great if that was how it always worked. It is true that some, > most, modern browsers will attempt to try the second/n address after > attempting to connect for a few seconds. Unfortunately there are other > issues like DNS caching, proxies, mobile browser etc that either don't > support or disrupt the IP failover. In these cases it will resolve to the > first IP, if that fails then it will just error out.
AFAIK, browsers just make a sys call to OS-level DNS resolver, so that does not really depends on how modern your browser is. Multiple A records is a core part of DNS. Caching DNS servers must respect TTL values and refresh cache accordingly. > > The round-robin does help with failover, but is more to distribute load to > clustered applications. This is part of the reason for floating IP's. If a > server fails then a different server can take over the IP and inherit the > load that would be sent to the server from the round robin. > > That being said, I was hoping that people who use scalr had best practices > or possibly Scalr has some IP take over "magic" for servers that go down. I > know that an Elastic IP can be set to a server role and then reassigned to a > new instance that is replacing a failed one. This might be good enough if it > comes up fast enough... just looking for some tips. Instance starts to respond on elastic IP after 30 seconds to 5 minutes after it was assigned. This is a key problem. --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "scalr-discuss" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/scalr-discuss?hl=en -~----------~----~----~----~------~----~------~--~---
