News for hackers. Stuff that matters.
Re:Sled: Embedded Database Written in Rust by cft on Friday February 21, 2020 @12:16AM (#22381084)
Not sure if this this still relevant, but sled had a memory leak that precluded us from using it for an actual production project 3 months ago:

https://github.com/panicfarm/sltest

Re:Sled: Embedded Database Written in Rust by krenoten on Friday February 21, 2020 @03:04AM (#22381625)
On sled 0.31 the memory growth does not run away as with the earlier version. Note that some in-memory metadata tracking does accrue as nodes split and sled tracks where to find them in the future when the actual data is paged in. But this is still a fraction of data stored.

The important thing is that sled is not stable yet. Don't bet your business on databases less than 5 years old. Sled doesn't turn 5 for a few more months, and I'm ironing out a few issues still that relate to production readiness. For now, treat it as a cache, as a responsible SRE would treat any database younger than 5 years old.

Re:Sled: Embedded Database Written in Rust by shakna on Friday February 21, 2020 @01:55AM (#22381401)
The reproducible bug is great, and I went looking to see if it might have been fixed, but I didn't find the same bug in the list.

Did you let sled know about the bug? Can you point me in that direction?

Re:Sled: Embedded Database Written in Rust by cft on Friday February 21, 2020 @03:20AM (#22381689)
Yes, I wrote in sled's discord on Nov 14 and got this from the author:

hey @panicfarm, thanks for the reproduction code! I think this is the same issue that kerugma ran into and mentioned a couple days ago. I'm curious if it still happens on sled master? I just did a huge refactor and sometimes a bunch of issues just shake out. Anyway, I will turn this into a proper regression test and get the page_out method of PageCache to properly be called, which is the underlying root.