Which NAS do you use with your Hackintosh ?

[trs96]

Try Synology DSM. I've never needed to SSH in there to do any repairs. It "just works".

Must be a reason these are the Amazon best sellers. If they were making junk they wouldn't sell.

 

[zipb]

Try Synology DSM. I've never needed to SSH in there to do any repairs. It "just works".

I never had a problem with filenames either... But I still connect to the NAS via AFS. Not sure if filenames are a problem with NFS or CIFS...

I haven't had any issues with Time Machine since I started using my NAS for it in May.

I had all sorts of disconnect problems with DAS over USB and later with Thunderbolt to SATA.

Really? Trouble with a USB drive connected straight to your Hack and formatted as HFS+?

 

[zipb]

Must be a reason these are the Amazon best sellers. If they were making junk they wouldn't sell.

View attachment 555027

My 2c:

If you just want to store movies and mp3 files for your entertainment, a consumer NAS may be a solution. I use a RPI & WD USB drive for this. Cheap and cheerful. I have a backup of the USB drive, BTW. Stored somewhere else.

But for mission-critical stuff, you need pro hardware, a service contract and trained staff to do it properly. And a backup strategy, of course.(for example, veeam)

I will not put my family photo's on a consumer NAS only. These are on a SSD in my Hack, backed up by a USB drive, plus Adobe online storage,plus Backblaze.

Important audio projects, same. 1 local copy, 2 in the cloud.

 

[pastrychef]

Really? Trouble with a USB drive connected straight to your Hack and formatted as HFS+?

Yes. On large copies, I've had them disconnect. Also, with USB drives connected, I always get "not ejected properly" notifications when waking from sleep.

 

[zipb]

Yes. On large copies, I've had them disconnect. Also, with USB drives connected, I always get "not ejected properly" notifications when waking from sleep.

Weird. I've seen the "not ejected properly" messages too, but these were fixed by lowering RAM speed. Yes, really. Lowering RAM speed...

Copying lots of stuff over USB2/3.x is not a problem for me. Just works. The only problem is that my USB cradle will sometimes attach to USB2(slow) instead of USB3 if I switch it on while the Hack is running. Replugging the USB cable or on/off switching fixes this.

 

[pastrychef]

Weird. I've seen the "not ejected properly" messages too, but these were fixed by lowering RAM speed. Yes, really. Lowering RAM speed...

Copying lots of stuff over USB2/3.x is not a problem for me. Just works. The only problem is that my USB cradle will sometimes attach to USB2(slow) instead of USB3 if I switch it on while the Hack is running. Replugging the USB cable or on/off switching fixes this.

Adjusting RAM speed is not an option on a real Mac and that's where I've experienced the "not ejected properly" notifications.

My guess on why the connections break on large transfers is that something on the USB drives overheat. I've experienced this on multiple USB drives/enclosures.

My NAS has been, by far, the most reliable. It runs 24/7 and never complains.

 

[cacamaca]

I have been using Xpenology since DSM 6.0 came out with 0 issues. I have 4 Xpenology servers plus 2 Synology machines - 216j and 918+. In addition I am running ESXi 7.0 on a Mac Mini Late 2012 with DSM 7.1, DSM 6.2.3, Zevenet load balancer and PhotonOS VMs (I really like PhotonOS for Docker containers).

On DSM I run TimeMachine, Plex, Bitwarden server, WordPress and all the services that have been automated to populate Plex library automatically.

The reason for the number of Xpenology servers (all running DSM 6.2.3 for now) is the 3-2-1 backup rule plus Synology video surveillance.

 

[nextsak]

Synology DS418, here!

 

[c-o-pr]

In the meantime I managed to finally upgrade the PSU in the Jonsbo N1 NAS box. I got hold of a Silverstone SX700 which is a decent SFX PSU that provides up to 700W power (which is enough to provide power to all 7 drives in the system which I plan to eventually have).

View attachment 554731

I've run into issues with macOS getting wonky with external NVMe after an incremental clone.

Then it got weird with an internal clone, where the access grinds to a halt.

By sheer luck of goofing around, I connected the problem to APFS snapshots on the clone target.

I suspect there's a macOS bug with APFS snapshots.

What I'd been seeing if Clone targets with snapshots (CCC enables default in later releases) will unpredictably hang when incrementally updated.

And for external drive (e.g. USB) once the pathology sets the drive may hang on any write. Worse, it may corrupt the clone if you try deleting its snapshots.

Try disabling APFS snapshots on the clone target before making any clone.

Source snapshots are ok.

For internal targets, the pathology is less severe, but incremental clone operations may slow to a crawl. Disabling and deleting snapshots on internal target can overcome the hangup.

This story is based on my direct experience from a few days ago.

* * *

Regarding 700W for 7 drives that need 7-10W each, you've mosdef got enough power!

I've got 14 spinners in a box with with a 5Ghz all-core 11900K running all out inc a 6600XT plus trimmings all on a Corsair RM650x with no power issues. So i figure your 7 drives are well covered.

 

[Middleman]

So an update on my Jonsbo N1 box. I've fully populated the drives with the GA-Z170N-WiFi and set up my first RAID-Z setup. With 5 x 2TB drives on a one drive failure configuration and the TrueNas boot drive installed on an Intel 256GB SSD it gives me just over 7.13TB of storage space. I've created a storage pool and set it to run both SMB and AFP protocols. So far it's giving some decent performance from the network: