We have seen people wanting to have something like ZooKeeper without the
reliability of permanent storage and are willing to work with loosened
guarantees of current Zookeeper. What you mention on log files is certainly
a valid use case.
It would be great to see how much throughput you will be able to get in such
a scenario wherein we never log onto a permanent store. Do you want to try
this out and see what kind of throughput difference you can get?
On 7/19/10 8:35 PM, "Ashwin Jayaprakash" <ashwin.jayaprak...@gmail.com>
> Cool. I've only tried the single node server so far. I didn't know it could
> sync from other senior servers.
> Server/Cluster addresses: I read somewhere in the docs/todo list that the
> bootstrap server list for the clients should be the same. So, what happens
> when a new replacement server has to be brought in on a different
> IP/hostname? Do the older clients autodetect the new server or is this even
> supported? I suppose not.
> Log files: I have absolutely no confusion between ZK and databases (very
> tempting tho'), but running ZK servers without log files does not seem
> unusual. Especially since you said new servers can sync directly from senior
> servers without relying on log files. In that case, I'm curious to see what
> happens if you just redirect log files to /dev/null. Anyone tried this?
> Ashwin Jayaprakash.