[email protected] writes:
> We're looking at this solution as part of a web app we're building, I
> wanted to ask the community what their experience is with benchmarks, how
> many reads/writes can this system manage at what point will it need to be
> scaled out.
I don't know of any.
> This brings me to my next question, I know Nikolaus has mentioned on a few
> occasions that this doesn't scale above one mount.s3ql for writes, although
> you can add a few extra read mounts through sshfs, as mentioned here, would
> this improve read performance or would all reads go still go through the
> one mount (ie creating a bottleneck and not improving performance...)
Everything would still go through the same mount. But that's also the
case for e.g. ext4 or btrfs, so it's probably nothing you'd have to be
concerned about.
That said, accessing S3QL (or any other file system) through sshfs is
certainly not going to make anything faster - quite the opposite.
> Thinking about clustering and performance it seems as though the system is
> built off the back of a sqlite db, would this be possible to separate from
> the system thereby making it possible to have multiple upload/download
> points all write to same db??
Everything is possible in principle, but I don't think anyone is
planning to implement this. It would most likely require switching
SQLite for a different database.
Best,
-Nikolaus
--
GPG encrypted emails preferred. Key id: 0xD113FCAC3C4E599F
Fingerprint: ED31 791B 2C5C 1613 AF38 8B8A D113 FCAC 3C4E 599F
»Time flies like an arrow, fruit flies like a Banana.«
--
You received this message because you are subscribed to the Google Groups
"s3ql" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
For more options, visit https://groups.google.com/d/optout.
