Hacker News Reader

Comment by CJefferson | Hacker News Reader

I couldn't see any fsync (or similar) in your filesystem code (sorry if I missed them) -- it doesn't seem like a completely fair comparison, as sqlite has various promises about commits being completely made, etc, which you won't get from just chucking files on a filesystem.

View on HN

Yeah. I mentioned that in the final test where I write to a temporary file, then rename. This is much slower, probably due to an implicit fsync.

You can disable the fsync calls in sqlite if you wanted to do a little better with this benchmark. You're also explicitly choosing the WAL in your go-sqlite3 configuration which is not at all replicated by your filesystem test. I think, honestly, that you're just going to mislead and confuse people who don't know any better with this writeup. I can write faster to /dev/null, too, but that isn't a very interesting comparison.

Does /dev/null support sharding?

Yes.

Even better, it's embarrassingly parallel!

I am not aware of implicit fsyncs. Can you please link to what you are referring to?

Probably in reference to Ext4's[0] configurable rename-replace behavior under auto_da_alloc

[0]: https://www.kernel.org/doc/Documentation/filesystems/ext4.tx...

Edit: note this isn't a universal property either, so it's still wrong: https://btrfs.wiki.kernel.org/index.php/FAQ#What_are_the_cra...

Yeah considering that the test is on btrfs, I would really be interested in knowing what OP is referring to.

Also, even if we were on ext4, it's worth pointing out that the ext4 documentation refers to apps that rely on that behavior as "broken applications", so it's hardly a good term of comparison.

There's some 10-year old webpage that used to be popular on HN that goes into data integrity issues and fsync and filesystem databases that someone should dig up so you can read.

https://danluu.com/file-consistency/, https://stackoverflow.com/a/41362774

Yes that's the one.

And damn near every application is "broken" according to the ext4 documentation.

If you only care about storing data and not doing any complex operations or query against it why are you using something like sqlite as well?

It’s not a comparison as being in sqlite makes the ability to access this data significantly easier. This is comparing apples and dogs and i don’t see the merits.

It's not uncommon to see people advocate for using sqlite as an alternative to flat files or json blobs. In particular, transactions provide nice properties vs a web of separate flat files, and the stronger schema can be a good alternative to json. There's a lot of great existing sqlite tooling, too.

Not only other people, the creators also advocate for it as a file format:

https://www.sqlite.org/aff_short.html

https://www.sqlite.org/appfileformat.html

Sqlite archive files are somewhat interesting too: https://www.sqlite.org/sqlar.html

The history around mongrel2 using SQLite as a config file format may be instructive. Seems like there was a lot of resistance in that case.

In one application I use sqlite to store JPEG thumbnails. That's it; there's nothing else in there. It's super handy; I specifically needed to reduce the number of files that I open and close because that's slow on Windows/NTFS. SQLite made this trivial. I could have managed a binary pack format on my own but I didn't have to.

You could have easily done that with a single file with a tail header for look ups.

Easier than SQLite? With less bugs and better tests?

SQLite advertises itself as an fopen replacement. Sounds like a perfect match for parent’s use case.

Depends on the indexing needs, right? Something like IFF should be fine if you need a sequence of images to store that you don’t need to actually query for individually.

Always, but I find it really hard to see how not using SQLite in this use case is the better option. Maybe if you need every single percentile of performance? In 99% of other cases just go with SQLite.

Yes easier, as in time to integrate and get working correctly. I’ve used SQLite extensively over the last 10 years and yes it’s a good solution, but not a replacement for fopen. A flat file would be easier to integrate, test, and harden over SQLite. Would take about the same time as integrating SQLite into a system for the same purpose and would be easily extendable to support features as the system grows.

> Would take about the same time as integrating SQLite into a system for the same purpose and would be easily extendable to support features as the system grows.

I find it much easier to add features to my post-2007 projects (when I started using SQLite) for the specific reason that I can open the SQLite file in a GUI and pretty quickly see what’s going on with data organization (schema) and how the customer uses the software I wrote (ie by what columns they actually use/misuse).

Prior to that, there’s various versions of my b-tree library and lots of zips, or linear text indexes, or any combination of whatever fit the need. Data storage implementation needs to be reasoned about in detail for each pre-2007 revisit in ways that don’t happen with SQLite projects.

What's the UI you use for viewing SQLite files?

At client site to debug issues and deal with things like bulk product name changes https://sqlitebrowser.org/

For more involved work that I do at my shop or on my laptop, DataGrip by Jet Brains is great. Before I got fed up with Apple, I used https://menial.co.uk/base/

DataGrip’s benefit to me is mainly reworking a customer DB at the ER diagram level, then I manually update my code to match.

By the way, SQLite started out as a TCL extension. It works really well in TCL and good old TCL/TK/SQLite gets CRUD done fast. I almost feel bad looking so heroic.

I'm aware; I mentioned this at the end of my post. This is still easier. I want to clearly express how absolutely trivial it was to use SQLite for this. I didn't really have to think about anything.

Another alternative that I've seen used is a zip or tar with no compression if you are just appending files and reading but only rarely updating or deleting.

But sqlite is still better, it is more reliable, a bad write on that end of zip index destroys the whole zip archive and sqlite also gives you a lot more potential flexibility later on if you need to add metadata or something else. It is better in terms of inserts and deletes, although you will still need to vacuum.

A bad write on a zip file destroys the O(1) seek time, but it doesn't destroy the zip. That goes back to PKZip trying to work on floppies and over modems. You can still do an O(n) seek on a particular file, or expand and recompress the file to recover whatever isn't truncated.

For this situation it does matter, but it is recoverable.

Oh interesting! Good to know

That property came back into use when we crossed the 2GB and then 4GB threshold for archives, because the TOC uses a 32 bit integer to reference backward into the file. A lot of implementations used a signed int, and even when they fixed it, organic growth in bandwidth and thus content size ran you into the wall in another couple of years. We had collectively mostly supported UTF-8 filenames around that same time period, so it was going from one problem to the next (or choosing between implementations that had fixed one but not the other).

For the 32 bit address problem, you could read forward from the front, and as long as no entry was over 4GB, you could still read the file. If the file count was low enough you could cache all of the entry objects, and if reads dominated opens, then you were functionally back to O(1) access (but O(n) startup).

There was a 64 bit extension going around, but when I stopped working with zip files every day I stopped tracking the progress.

Indeed, "SQLite does not compete with client/server databases. SQLite competes with fopen()."

https://www.sqlite.org/whentouse.html

SQLite makes seeking in a file using a lookup table look like a complex operation from the developer’s perspective, which is absolutely a good thing unless you have a strong alternative rationale for your use case.

Reply on news.ycombinator.com