 Tom here from Learn Systems, and we did some benchmarks here on May 20th of 2022, also same date I'm publishing this video, on TrueNAS Core 12.0 U8.1, then we upgraded in place to TrueNAS Core 13, then we moved it to TrueNAS Scale 22.02.1. This was all done on exactly the same system, and each time we ran the same series of benchmarks using Pheronix, so we could get a baseline and answer the question, as of today, and as of those versions of the software, is the performance difference between each one of these versions of TrueNAS? I think this is an important question because a lot of what we do when it comes to consulting is going to be around using it as storage target for a hypervisor. I know that's not the entire equation because TrueNAS Scale, billing built on Debian, well, has lots of great features such as all the different Docker support and a large catalog of applications that can be on there. And if your use case isn't the absolute best and most performance, but more about the application support that may not even exist for you on TrueNAS Core, then Scale is still a good choice. So I'm not here to tell you, you shouldn't use Scale. I'm here to discuss the benchmarks of these results, so you have an idea of whether or not Scale is a fit for what you need to do. And if what you need to do is those applications, that may matter more than a performance is the bottom line. But before we dive into the details of this video, let's first. Are you an individual or company looking for support on a network engineering, storage, or virtualization project? Is your company or internal IT team looking for someone to proactively monitor your system security or offer strategic guidance to keep your IT systems operating smoothly? Not only would we love to help consulting your project, we also offer fully managed or co-managed IT service plans for businesses in need of IT administration or IT teams in need of additional support. With our expert install team, we can also assist you with all of your structured cabling and Wi-Fi planning projects. If any of this piques your interest, fill out our Hire Us form at laurancesystems.com, so we can start crafting a solution that works for you. If you're not interested in hiring us but you're looking for other ways you want to support this channel, there's affiliate links down below to get you deals and discounts on products and services we talk about on this channel. And now back to our content. Now let's first talk about the test scenario before we get to the results. And each time we use exactly the same system. As I said, we started it with TrueNAS Core 12, migrated it to 13, and then migrated it over here to TrueNAS scale. It's just an AMD Ryzen system that we have in our lab. It has 29.3 gigs of RAM. That's because it's got a shared video. So it's not quite 32 gigs of RAM. And you can even go here like the system settings and look at the boot. And you can see the previous version where we had each one on here, even the 13 release. But through each one of these tests, we always used exactly this hardware. And we didn't reconfigure the pool each time because there wasn't a need to. When you do the in-place upgrades, you're able to keep using the same shares and the same pools. And all this was just done over NFS. I didn't have time to do an additional Ice Cuzzy test as well. But NFS was good enough to give us an idea of the performance. We also inside here, we're going to go to the storage, and we're going to look at the ZenLab test here and look at the settings. And you'll see that we have the sync disabled. So that was done with all of these tests. I didn't go into deeper scenarios of testing with different log drives or different BDEV configurations. We kept this really simple with just three SSDs because we've done other testing and notice performance differences, but we wanted to narrow it down to just kind of a basic setup and see where it was at. Now let's break down the test results. We use Pheronix with the flexible IOTester as you can see is version 3.29. All this is linked down below in description so you can cover it and dive into a little bit closer. Each one of these tests, like random read, then it has a different block size. We didn't want to just test it with a single block size or an average. We test it with a couple different block sizes. The way we did the testing though, first we ran all these tests on the VM that was running on our hypervisor using TrueNAS as a storage target via NFS. And we did the in-place upgrade. So we started at 12. After we did the test on 12, we moved it to 13, but didn't update the feature flags. Then we updated the feature flags on the ZFS pool because this was a point of contention some people have when I ran the benchmarks saying, well, if you don't update it, maybe you missed out on some of the performance. Not much of a difference between before and after upgrading the feature flags. Then we moved it over to TrueNAS scale, which has the same new version of ZFS, so there's no feature flag changed there and ran the same benchmarks. And on this random read, one big block, your megabytes and your IOPS pretty much, they're all neck and neck with each other, almost within margin of error for the performance. Not much of a difference. But we went to the 4K block size random read, we got about half the performance out of TrueNAS scale. Pretty substantial drop in performance on that. And oddly, TrueNAS Core 13 was just slightly less performant on that particular test. Scrolling down further back to random write, one meg target, TrueNAS core just a little bit slower. It's going down a little further. And we have the performance difference where it gets really bad. Random write and random write with 4K blocks, TrueNAS 13 actually outperformed TrueNAS Core 12 when it came to the random 4K block writes, but TrueNAS scale is not even on the same category. It is just absolutely, I mean, at least it was half on the other one. This is just really bad performance it had when it came to doing random writes. Scroll down a little further. Now we have random reads with 128K block sizes. They're all neck and neck with each other again. Then we go back to random write with 128K block sizes. And we're about half the performance with TrueNAS scale. So it's kind of back and forth with depending on the workloads that you throw at it. Random writes here 128K still about half the performance. Then we go over to sequential reads, the large one meg blocks, all neck and neck. So they all handle the one meg block fine. We do sequential reads, 4K block, not quite half, but performed a little bit better with sequential read here with TrueNAS scale. Scroll down a little further. And then we can do sequential write with the one meg block, still a little bit less performance, but not bad. TrueNAS core though, coming out ahead on that one. Scroll down a little further, sequential writes. And they actually handled the writes on the one meg block still reasonably well. Core 13, I'm not sure why, but before doing the upgrade, we got a little bit more performance out of it. Almost seems like a fluke, but we ran the test and let Veronica do it and do the margins of error. And this is what it came up with. This is, you know, all repeated tests here. But sequential write engine, 4K block, once again, TrueNAS scale, not handling those 4K block writes very well. Scroll down a little further. Here's your IOPS, same results there for those 4K blocks, sequential read, sequential read. And you go through and get here to the bottom. And once again, the 128K blocks, it's still performed reasonably well. So there you have it. Those are the results as of May 20th, 2022. TrueNAS core 12 still performs well, 13 edges it out in a couple small spots. Maybe TrueNAS core was a little bit faster, but TrueNAS core 12 is a, you know, really enterprise stable 13. As you know, for those of you that keep asking about this, it's said community stable, but not enterprise stable. They consider U1 when it becomes an enterprise stable suggestion by U1. They could add even more performance tuning. TrueNAS scale, still a great product. It's amazing. They've gotten all that tooling done as fast as they did to, you know, do a complete operating system swap, still have ZFS as the base, still have a common looking interface to the way TrueNAS core works. But obviously not there yet in performance, but performance isn't everything. The application catalog that comes to TrueNAS scale makes it a very compelling storage product. The clustered storage is really important, but for people that need lots of, you know, 4K writes at high performance because of a database workload or something they have shared over NFS or using as a storage target, it's not the right fit for them. It would be better to stay with core in that case. And the fact that they have an easy migration to go from core 13 in place over to scale, it even sets up all your shares, keeps all your data in place, shows you have a backup before you do these things. But the process that we went through was relatively easy migrating between three different versions of TrueNAS to do these results. So I'm impressed with how well all of that worked. Hats off to the AIX systems team for all the tooling involved. I don't want this to be your final decision on which TrueNAS is right for you, but it should serve as a data point and see if that's a data point that is relevant to you. And that is down below the link depth so you can stare at the numbers, scratch your head over them, or, hey, Fronix is free. You can run all your own testing and come to your own conclusions. I just did this because a lot of people asked the question. We do too. We were curious where it was at. And we didn't want to just keep those numbers to ourselves. We want to share it with you. Leave your comments down below or head over to my forums for more in-depth discussion. Let me know what else you want to see benchmarked. And if we have some time, maybe we'll do some more benchmarks. Alright, thanks. And thank you for making it all the way to the end of this video. If you've enjoyed the content, please give us a thumbs up. If you would like to see more content from this channel, hit the subscribe button and the bell icon. If you like to hire a short project, head over to lauranceystems.com and click the hires button right at the top. To help this channel out in other ways, there's a join button here for YouTube and a Patreon page where your support is greatly appreciated. For deals, discounts, and offers, check out our affiliate links in the description of all of our videos, including a link to our shirt store where we have a wide variety of shirts that we sell and designs come out well randomly. So check back frequently. And finally, our forums. Forums.LauranceSystems.com is where you can have a more in-depth discussion about this video and other tech topics covered on this channel. Thanks again for watching and look forward to hearing from you.