DISCLOSURE: After chasing both Synology and QNAP for a couple of years now to try out their various Home NAS products in exchange for a full blog review, only to be met with “We’ll give you 10% off,” I bit the bullet and did it myself, integrity intact. Here is my full review of my first two weeks with a Synology DS1813+. Synology or no member of their team has approached me about writing this, and in no way am I being compensated for it. This should in no way be viewed as a competitive analysis, but simply one enthusiast’s simplification of his home media and data management.
All links in this article are Amazon Affiliate links. Please consider using them to make any purchases if you choose to go down the same path. All proceeds go towards the continued support and maintenance of DatacenterDude.com. Thank you!
So how did I get here? Why would anyone go down this route for home use? Ruling out the obvious demo/lab use-cases, which I have for the purpose of this article, it brings a TON of functionality to the table, which will go largely unused by most end-users. I was a long time FreeNAS user, but the frequency of updates, lack of confidence in open source housing all of my personal data, and the amount of heat and noise that rig put out, pushed me back to using external USB drives. And it simply got out of hand. Before we get too deep into the philosophy, let’s lay out my home environment first so we can really put things into context.
I have a MacBook Pro (2011) that is my “main” machine. I use it for work, run 10-12 various VM’s in Fusion, etc, but also use it for Final Cut Pro X for some pretty intense video editing, as well as GarageBand and Skype for recording multiple podcasts. With 16GB RAM, an i7, and a 512GB SSD, it is a workhorse, and short of getting a Mac Pro for $5k+, likely won’t be eclipsed very soon. Combine this with (2) Apple Thunderbolt displays, a 12-channel mixer, and a couple of Heil microphones, and you’ve got the perfect workstation. At least I think so.
Attached to this, I had (2) 1TB USB3 drives for live storage, (1) 2TB USB3 drive for Time Machine, and most recently acquired a 2TB WD Velociraptor thunderbolt rig (RAID-0) because I was having performance issues using the USB drives. The Thunderbolt performance is magical, for what it’s worth, but that’s a whole other topic.
I also have a Mac Mini, that essentially sits there and runs iTunes. That’s about its only purpose in life. It serves my iDevices and Apple TV in the living room, to which I have built quite the library (1.5TB of ripped bluray movies, 100k+ songs, photos backups, etc) and storing all of THIS was a Seagate 4TB USB external drive, which was also being used as the Time Machine target for that device (bad idea, I know).
My day-to-day data mgmt had become a chore. And something that was really beginning to have to be “managed.” I tried going down the path of CrashPlan initially, but it didn’t become very cost effective over time, and the up-front legwork involved was more than I was willing to go through.
“OK, now what?”
Everyone recommends the Synology’s. Seriously. EVERYONE. So I started the month-long research effort. How many bays did I need? What sort of drives are people using? Can I get away with a 5-bay, or should I go with the 8-bay so I’m not pissed-off at myself a year from now when I need to upgrade anyway? How do the expansion units work? Is eSATA really good enough to expand a volume across enclosures? Etc etc etc.
The other question that came into play was networking. To-date, I had survived on an unmanaged 8-port Cisco/Linksys gigE switch. But the Syn had 4 gigE ports, so getting it without being able to do LAGG would be a bit silly.
Ultimately, after much back and forth, including swallowing hard at the price (~$2k), here’s what I ended up with:
….pushing the “Place Order” button on Amazon was literally a 2-hour endeavor of sitting there staring at it. I knew what I needed to know, the questions had been answered, I was sold on the technology. “Yea, but TWO GRAND?!” At the end of the day, I’m a firm believer in doing things right the first time, and not half-assing it.
Since I have Amazon Prime, it all got here within the next couple of days. It took 5 minutes to unpack the Synology itself, and about an hour to unpack all 8 of the individually boxed and packaged WD drives. About an hour of configuring the Volume and initializing (8) 2TB drives with an SHR-2 raid array
(Synology’s proprietary RAID-6 spinoff) and she was up and running. The UI is beautiful. They’ve taken all of the best parts of the Mac and Linux UI, and made DSM into this amazing mix of the two. This netted me ~12TB usable with the SHR-2, versus ~10TB usable with Raid-6.
UPDATE: Synology reached out to clarify that the SHR-2 offering is a linux-based offering, and not proprietary to Synology. Here is a link to their wiki, explaining how it all works. And here is a link to a support KB/FAQ describing the restore process should the worst-case scenario happen with a DiskStation failure.
OK, now we need some LAGG. Can’t rave enough about the Cisco SG300. It’s a fanless, quiet-as-a-mouse, managed 10-port gigE switch
, and even has a couple of gbic ports in case I ever wanted 10GbE. (SFP’s are 1GbE only). There was a bit of a learning curve, as it operates a lot like IOS in enterprise Cisco switches. I was reminded the hard way that even through the GUI you had to ‘wr mem’ or Save your config, as a reboot wiped mine out several times before I did the /foreheadslap and realized what was happening. Getting LAGG setup was also tricky and I actually had to RTFM because there is a certain order of operations between enabling LACP and adding the appropriate port profiles to the LAGG. Once done, though, I expect to quite frankly never have to touch it again. And I haven’t since day 1.
Storage? Check. 4-way LAGG networking? Check.
Let’s Begin. Again.
Now to create some shares and let the data consolidation project begin.
All-told, I’ve got a sum total of about 2.5TB of data. This was being moved in from all of the various USB drives I mentioned above, and took a couple of days to get moved over. I then spent the better part of a weekend getting digitally organized. I think this is where most people (including a lot of my friends) fall down. Hard. On their faces. And then they bitch when their hard drive fails or computer breaks and wonder why they can only access a backup from 6 months ago. It’s a lesson that a lot of us, including myself, have learned hard. This was a theme for this whole project, but we’ll get back to that when we talk about Backup.
Getting digitally organized involved consolidating everything into one “share” on the Synology. I actually ended up deleting roughly 1TB worth of duplicate data, where I had multiple copies of things saved, putting my trust into the double-disk protection of the Synology. Fitting enough, I named the first share “NAS.” Inside of the NAS share, there are an array of folders and subfolders containing everything under the sun (I finally deleted a lot of my Microsoft ISO downloads from my technet. Something tells me I’m going to regret that, but I digress…).
This is something that requires ongoing grooming. Even as I write this, I’m copying another 170GB of recent tape capture footage from the thunderbolt drive over to the Synology.
Now that the data is consolidated, and somewhat organized, let’s get into some of the nitty gritty details of how I’m using it in each of the various ways.
Double-Disk Parity Storage
As a NetApp’er that bleeds NetApp Blue, I wholeheartedly believe in RAID6-DDP-style deployments, such as our RAID-DP. To be honest, I didn’t buy this for performance. I bought it more for that “Sleep well at night knowing my data is safe,” but it’s actually quite performant. For comparison, here are the three drive types, USB, Synology NAS, and WD Velociraptor Thunderbolt.
With this setup, I know that I can lose two drives and my data will still be intact. Something that was missing with the array of various USB drives I had plugged in. This was arguably the biggest checkbox on the list of reasons to get a big home NAS. Mission accomplished.
This one is touchy. One of the biggest gripes to Apple in recent years is the inability to serve an iTunes library without a running instance of iTunes. Unfortunately, that hasn’t changed. Synology did however come up with a solution that involves some free iDevice Apps and leveraging AirPlay. As much as they advertise that they can serve a library to an AppleTV or an iDevice, understand that it’s not native/headless, and is proxying via AirPlay through an iDevice. Not their fault. Nothing they could do about it really. C’mon Apple.
My preferred solution stays the same. I already had a Mac Mini running an iTunes library and it’s worked brilliantly for me for 3+ years now. It was simply a matter of moving it from the aforementioned 4TB external drive to the Synology, mounting the NAS share, Option-click iTunes, and point to the iTunes Library.itl file, and all was back to normal. I will say that Apple has done a much improved job of making the entire library, including metadata, more portable. But copying a 1.5TB iTunes Library takes a few days.
Lots of people also use PLEX as the platform-of-choice for media sharing, but considering how in-bed I already am with Apple, I prefer to stay within the ecosystem. Once you get the process down using MP4Tools and iDentify to populate the library, it’s really simple. If anyone has a specific request of using iTunes exclusively, converting AVI/MKV files, auto-tagging artwork and IMDB metadata, and adding to iTunes, I can certainly write that up in a separate post.
UPDATE: Many have sent in requests for a thorough explanation of my iTunes Library management. I’ll see if I can get that up this weekend.
UPDATE 2: After all of the feedback and requests, here is a detailed walkthrough of my iTunes configuration, workflow, and metadata management.
For now, I still use iPhoto for managing everything. It has done everything I need it to do, from managing collections and album, to providing “just enough” retouch and editing tools for the hobbyist photographer category I fall into. The way iPhoto manages photos can be a little bit daunting without taking time to truly understand the database and revisions logic and tracking that it does. All iPhoto content is stored in a single “package” file, not unlike a lot of Apple applications, making it easy to move things around if need be.
For performance purposes, I have opted to leave the iPhoto Library on my Thunderbolt drive. If you need justification as to “why,” refer to the speed test results above. But, that drive and it’s contents are included in the Time Machine backup profile for the MacBook Pro.
On the back of the Synology are (2) USB-3 ports. Now that I had a bunch of extra USB-3 drives laying around, I didn’t want to consume expensive 2TB WD Red disks for backup copies. (Sound familiar, IT folks?) One of the coolest features I’ve used of the Synology so far is to plug in a USB3 device into the back, and have it advertise on your network as a Time Machine target.
Now, both my MacBook Pro, as well as my MacMini can backup with Time Machine, and leverage sparse bundles stored on an ext4-formatted USB3 disk. It does it all for you. Easy-Peezee. I would also recommend making sure that “Local Snapshots” are disabled for Time Machine, especially if you’re using an SSD locally. This can be accomplished by running:
sudo tmutil disablelocal
…at a Terminal prompt.
So, if we review, I’ve got all of my data stored either on the Synology with double-disk protection or on a Thunderbolt DAS being Time Machine’d every hour. Brilliant.
Now let’s talk about arguably the most controversial topic…
Backup the Backup
What kind of storage guy would I be if I said this was “good enough?!” Frankly, it really is. For 99% of the world, this is a 1000x improvement over what you’re likely [not] doing, which for a lot of people includes no backup at all aside from what you’re sync’ing with Dropbox. Sound familiar?
I’ve said for a long time, “Backups are useless if they can’t be restored.” Living in Southern California, we’re all waiting for the next 7.0+ 1994 disaster again. Buildings crumble. Datacenter and CoLo facilities here in L.A. have “Earthquake sustainability” ratings. Most claim to be able to “handle” up to an 8.0. But I guarantee you your house or apartment building that was built in the 1950’s cannot (mine included).
So while the Synology is a 10x improvement for me over what I was using before, it still needs to get backed up somewhere else. One of the cool parts about the Synology is that it has built-in support for Amazon Glacier, as well as a few other cloud services. I’ve tried Glacier in the past and just wasn’t overjoyed about it. It’s not terrible, but it does what it says on the tin. It’s essentially meant to be cold storage. Hard archiving of data you won’t access very often. This is why it is so cheap.
As an alternative, I’ve heard nothing but rave reviews of late from CrashPlan users. I was excited when I discovered that some smart guy figured out how to install both an embedded java runtime and a headless CrashPlan server onto Synology diskstations. You also still run the local client GUI on your workstation and simply point it to the server running on the Synology. The HOWTO deserves its own writeup/post, but for those anxious to see how it works, you can grab the instructions here.
I’ve got about 30 days left in my initial seed of ~1TB. I’m aware they had a preload option, I’m just not willing to pay $125 to ship a drive back and forth. And no, you can’t backup to my Synology, CrashPlan users. :)
This is an amazing rig. I’m using 1-5% of what it’s truly capable of. I’ve had it running for 2 weeks (1 reboot 2 days in for DSM upgrade) and I actually thought I would be able to push this thing. What you’re seeing below is with a movie streaming to an AppleTV, CrashPlan upload/seeding, and Time Machine all running at the same time.
I have not even begun to play with all of the applications that it has in its own little “app store” that works a lot like a linux software repository. It also has a huge open source community with people hosting their own repositories, which the DiskStation fully supports adding. I’ve tinkered with the BT client, and while it’s fully capable, I still prefer using uTorrent on the Mac Mini. I’m sure I’ll switch over at some point.
If for any reason any of this has touched a vein with you, do not hesitate. Take the leap. Pull the trigger. Buy it. But don’t cut corners, and don’t half-ass cheap-out on it. Do it right the first time. You won’t be sorry.
Questions/Comments? Let me hear it! I’ll append the review with answers to the good ones!