cdown

This is actually something we're actively working on! Nhat Pham is working on a patch series called "virtual swap space" (https://lwn.net/Articles/1059201/) which decouples zswap from its backing store entirely. The goal is to consolidate on a single implementation with proper MM integration rather than maintaining two systems with very different failure modes. It should be out in the next few months, hopefully.

There are quite a few numbers in the article, although of course I'm happy to hear any more you'd like presented.

* A counterintuitive 25% reduction in disk writes at Instagram after enabling zswap

* Eventual ~5:1 compression ratio on Django workloads with zswap + zstd

* 20-30 minute OOM stalls at Cloudflare with the OOM killer never once firing under zram

The LRU inversion argument is just plain from the code presented and a logical consequence of how swap priority and zram's block device architecture interact, I'm not sure numbers would add much there.

Thank you for reading and your critique! What you're describing is definitely a real problem, but I'd challenge slightly and suggest the outcome is usually the inverse of what you might expect.

One of the counterintuitive things here is that _having_ disk swap can actually _decrease_ disk I/O. In fact this is so important to us on some storage tiers that it is essential to how we operate. Now, that sounds like patent nonsense, but hear me out :-)

With a zram-only setup, once zram is full, there is nowhere for anonymous pages to go. The kernel can't evict them to disk because there is no disk swap, so when it needs to free memory it has no choice but to reclaim file cache instead. If you don't allow the kernel to choose which page is colder across both anonymous and file-backed memory, and instead force it to only reclaim file caches, it is inevitable that you will eventually reclaim file caches that you actually needed to be resident to avoid disk activity, and those reads and writes hit the same slow DRAMless SSD you were trying to protect.

In the article I mentioned that in some cases enabling zswap reduced disk writes by up to 25% compared to having no swap at all. Of course, the exact numbers will vary across workloads, but the direction holds across most workloads that accumulate cold anonymous pages over time, and we've seen it hold on constrained environments like BMCs, servers, desktop, VR headsets, etc.

So, counter-intuitively, for your case, it may well be the case that zswap reduces disk I/O rather than increasing it with an appropriately sized swap device. If that's not the case that's exactly the kind of real-world data that helps us improve things on the mm side, and we'd love to hear about it :-)

Hey there! Article author here -- just found out this was posted here and going through the comments :-)

One of the earliest test versions of my patch actually did inode reuse using slabs just like you're suggesting, but there are a few practical issues:

1. Performance implications. We use tmpfs internally within the kernel in a lock-free manner as part of some latency-sensitive operations, and using slabs complicates that somewhat. The fact that we make use of tmpfs internally as the kernel makes this situation quite different than other filesystems.

2. Back when I was writing the patch, each memory cgroup had its own set of slabs, which greatly complicated being able to reuse inodes as slabs between different services (since they run in different memcgs).

After it became clear that slab recycling wouldn't work, I wrote a test patch that uses IDA instead, but I found that the performance implications were also not tenable. There are other alternative solutions but they increase code complexity/maintenance non-trivially and aren't really worth it.

A 64-bit per-superblock inode space resolves this issue without introducing any of these problems -- before you go through 2^64-1 inodes, you're going to have other practical constraints anyway, at least for the timebeing :-)

My e-mail address is already public from my kernel commits and upstream work. :-)

Hey! Facebook engineer here. If you have it, can you send me the User-Agent for these requests? That would definitely help speed up narrowing down what's happening here. If you can provide me the hostname being requested in the Host header, that would be great too.

I just sent you an e-mail, you can also reply to that instead if you prefer not to share those details here. :-)

I had no idea anyone would find this so fast, so I thought it would be safe to change the url to match the final title before posting anywhere -- how wrong I was.

It's back now. Thanks!

There are some systems which are memory-bound by nature, not as a consequence of poor optimisation, so it's not really as simple as "needed" or "not needed". As a basic example, in compression, more memory available means that we can use a larger window size, and therefore have the opportunity to achieve higher compression ratios. There are plenty of more complex examples -- a lot of mapreduce work can be made more efficient with more memory available, for example.

I mean, it really depends on your application how non-trivial the performance improvement will be, but this statement isn't theoretical -- memory bound systems are a major case where being able to transfer out cold pages to swap can be a big win. In such systems, having optimal efficiency is all about having this balancing act between overall memory use without causing excessive memory pressure -- swap can not only reduce pressure, but often is able to allow reclaiming enough pages that we can increase application performance when memory is the constraining factor.

how much time did I spend waiting to swap-in a page

You can do this with eBPF/BCC by using funclatency (https://github.com/iovisor/bcc/blob/master/tools/funclatency...) to trace swap related kernel calls. It depends on exactly what you want, but take a look at mm/swap.c and you'll probably find a function which results in the semantics you want.

Once desktop systems and applications support required APIs to handle saving state before being shut down.

SIGTERM and friends? :-)

If your application is just dropping state on the floor as a result of having an intentionally trappable signal being sent to it or its children, that seems like a bug.

I never saw programs I would actually want to use try to allocate that much

What about applications that have memory-bound performance characteristics? In these cases, saving a bit of memory often directly translates into throughput, which translates into $$$.

This isn't a theoretical, a bunch of services which I've run in the past and currently literally make more money because of swap. By using memory more efficiently and monitoring memory pressure metrics instead of just "freeness" (which is not really measurable anyway), we allow more efficient use of the machine overall.

One problem with relying on the OOM killer in general is that the OOM killer is only invoked in moments of extreme starvation of memory. We really have no ability currently in Linux (or basically any operating system using overcommit) to determine when we're truly "out of memory", so the main metric used is our success or failure to reclaim enough pages to meet new demands.

As for the analogy -- there are metrics you can use today to bat away grandma before she starts hoarding too much. We have metrics for how much grandma is putting in the house (memory.stat), at what rate we kick our own stuff out of the house just to appease grandma, but then we realise we removed stuff we actually need (memory.stat -> workingset_refault), and similar. Using this and Johannes' recent work on memdelay (see https://patchwork.kernel.org/patch/10027103/ for some recent discussion), it's possible to see memory pressure before it actually impacts the system and drives things into swap.

In my experience, a misbehaving linux system that's out of RAM and has swap to spare will be unusably slow.

Yeah, this is basically the main drawback of swap. I tried to address this somewhat in the article and the conclusion:

Swap can make a system slower to OOM kill, since it provides another, slower source of memory to thrash on in out of memory situations – the OOM killer is only used by the kernel as a last resort, after things have already become monumentally screwed. The solutions here depend on your system:

- You can opportunistically change the system workload depending on cgroup-local or global memory pressure. This prevents getting into these situations in the first place, but solid memory pressure metrics are lacking throughout the history of Unix. Hopefully this should be better soon with the addition of refault detection.

- You can bias reclaiming (and thus swapping) away from certain processes per-cgroup using memory.low, allowing you to protect critical daemons without disabling swap entirely.

Have a go setting a reasonable memory.low on applications that require low latency/high responsiveness and seeing what the results are -- in this case, that's probably Xorg, your WM, and dbus.

Here's the technical whitepaper: https://fbnewsroomus.files.wordpress.com/2016/07/secret_conv...

If you really want to store a "lifetime of photos", you should probably consider having a better long-term backup strategy than just putting it in locations that are managed by a single piece of software. :-)

Sure, some of these applications store a copy both online and on disk, but if your application chooses to run amok and delete everything, that's not really going to matter in many cases. I've seen horrifying things happen when iPhoto tries to sync with the cloud before. :-(