Unboxing 3 PETABYTES of storage!!

[nicklmg]

 

Buy Seagate Exos on Amazon (PAID LINK): https://geni.us/ZaXumr

[Tech_Dreamer]

FIRST!!!

[TempestCatto]

That's almost enough for 3 SlowMo Guys videos.

[Windows7ge]

I had a strong feeling they were going to go 45Drives again for this.

 

I do feel skeptical about their mention of a SLOG device. From my own experience with ZFS a SLOG device won't accelerate write operations with services such as SMB/CIFS or SSH/SFTP. Based on my own research a SLOG accelerates synchronous write which are often seen in VM applications and Databases. Meanwhile file shares are primarily asynchronous which means if the SLOG is used at all it will only be written to at the speed at which the pool itself can accept the data. This means the primary application of the box won't see any benefit from it.

 

I'm open to being told I'm incorrect but this was the answer I came to after a long time of researching SLOG devices. L2ARC should be fine though. ARC is data in ejectable memory though. If the system needs more for other processes it can dump the files. Based on the average size of there's though a L2ARC is probably not a bad idea.

[Alex Chipman]

I wish I could have Linus help Me with setting up a NAS.  My present NAS is a drobo i800 that gets read speeds of 250kbps and write speeds of 150kbps and when all 8 bays are full with 4tb hdds I only get 500GB useable space!

 

[GER_T4IGA]

2 minutes ago, Alex Chipman said:

I wish I could have Linus help Me with setting up a NAS.  My present NAS is a drobo i800 that gets read speeds of 250kbps and write speeds of 150kbps and when all 8 bays are full with 4tb hdds I only get 500GB useable space!

 

Probably horribly misconfigured. Ar you sure there isn't one 500GB drive left? Also if those drives are full then tehy will be slower.

[Alex Chipman]

Not to mention the fact that my desktop computer presently has 15+ external hard drives connected to it and I'd actually like to consolidate all of the data into 1 location.  Instead of me having to search through the drives (I have multiple drives for the same type of content) for a single file, not to mention the fact that whenever I do a backup I end up with having to take lots of time to deduplicate my backups and also my total 189TB of storage is full (500GB left).

[GDRRiley]

we going to see 2-4 systems given to other youtubers? sweet

 

I've been waiting for them to go epyc for years now. I joke every time they make a post. "It still isn't EPYC"

[Alex Chipman]

In my NAS I've stopped using that thing also since it took that NAS 45 minutes to boot up!

 

I build and repair computers for those who really can't afford new computers and I backup all of their data when O start working on their computers and sometimes I end up having to backup 1.5TB of their files.  And sometimes I have to work on multiple customer computers at one time.

 

14 hours ago, GER_T4IGA said:

Probably horribly misconfigured. Ar you sure there isn't one 500GB drive left? Also if those drives are full then tehy will be slower.

They were brand new 2TB ironwolf drives.  And after setting it up for dual disk redundancy that is all the useable space it would provide me is 500GB, which is worthless for a NAS, not to mention the fact that the only OS I could configure the NAS with is Windows XP via USB.  One Ethernet port had failed and the battery also had quit working.

 

Most of my money I receive (not very much) I spend fixing computers for those who have very little income as well.  In fact a lot of the time I end up "eating" the costs both of my diagnostic fees and parts and labor as well.  So having both a reliable network backbone and a better storage and backup solution are critical, just don't have the funds to upgrade the storage setup and also no funds for a new 2200VA UPS unit for my present server setup either.  My incoming funds are very unpredictable.  I don't even own a credit card either.

[unijab]

58 minutes ago, Windows7ge said:

ZFS a SLOG device won't accelerate write operations

 

just

 

zfs set sync=always dataset/name

 

 

[Windows7ge]

7 minutes ago, unijab said:

Just zfs set sync=always dataset/name

I tested that years back. Did not make a difference for write operations over SMB.

[RejZoR]

And I felt kinda cool when I stuck 8TB HDD into my system that already had 2TB SSD for combined 10TB of storage... LOL Then this guy stuffs bunch of spinny things into a single rack and can store entire PornHub on it. I'm now comforting myself with excuse that my system is quieter than the Storinator :P At least I win something ha! :D

[RastaGrzywa]

Linus, maybe You should consider tape drives for long term (archive) storage system?

[unijab]

58 minutes ago, Windows7ge said:

not make a difference for write operations over SMB.

 

I avoid windows, so i dont use smb

 

I use targetcli to export "luns" from my storage system to my proxmox system.

 

I have sync=always and a ssd drive for slog.

 

I max'd out ram on my motherboard so I dont have any l2arc configured.

 

heres a recent fio test from within a VM with a lun passed thru.

 

[root@cloud ~]# fio fiotest 
random-read: (g=0): rw=randread, bs=(R) 128KiB-128KiB, (W) 128KiB-128KiB, (T) 128KiB-128KiB, ioengine=libaio, iodepth=128
fio-3.7
Starting 1 process
random-read: Laying out IO file (1 file / 6144MiB)
Jobs: 1 (f=1): [r(1)][100.0%][r=1819MiB/s,w=0KiB/s][r=14.6k,w=0 IOPS][eta 00m:00s]
random-read: (groupid=0, jobs=1): err= 0: pid=25162: Tue Feb 18 21:21:34 2020
   read: IOPS=14.5k, BW=1807MiB/s (1895MB/s)(106GiB/60008msec)
    slat (usec): min=5, max=2299, avg=15.85, stdev=32.80
    clat (usec): min=992, max=196194, avg=8836.29, stdev=4824.39
     lat (usec): min=1005, max=196203, avg=8852.70, stdev=4824.48
    clat percentiles (msec):
     |  1.00th=[    3],  5.00th=[    4], 10.00th=[    4], 20.00th=[    5],
     | 30.00th=[    7], 40.00th=[    8], 50.00th=[    9], 60.00th=[   10],
     | 70.00th=[   11], 80.00th=[   13], 90.00th=[   14], 95.00th=[   15],
     | 99.00th=[   24], 99.50th=[   32], 99.90th=[   50], 99.95th=[   57],
     | 99.99th=[  127]
   bw (  MiB/s): min= 1317, max= 1823, per=99.99%, avg=1806.64, stdev=57.49, samples=120
   iops        : min=10540, max=14586, avg=14453.10, stdev=459.91, samples=120
  lat (usec)   : 1000=0.01%
  lat (msec)   : 2=0.76%, 4=11.95%, 10=49.58%, 20=36.30%, 50=1.32%
  lat (msec)   : 100=0.08%, 250=0.01%
  cpu          : usr=4.31%, sys=33.53%, ctx=248022, majf=0, minf=44
  IO depths    : 1=0.1%, 2=0.1%, 4=0.1%, 8=0.1%, 16=0.1%, 32=0.1%, >=64=100.0%
     submit    : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
     complete  : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.1%
     issued rwts: total=867358,0,0,0 short=0,0,0,0 dropped=0,0,0,0
     latency   : target=0, window=0, percentile=100.00%, depth=128

Run status group 0 (all jobs):
   READ: bw=1807MiB/s (1895MB/s), 1807MiB/s-1807MiB/s (1895MB/s-1895MB/s), io=106GiB (114GB), run=60008-60008msec

Disk stats (read/write):
    dm-2: ios=864084/2, merge=0/0, ticks=7528663/8, in_queue=7529636, util=99.88%, aggrios=867359/3, aggrmerge=0/0, aggrticks=7558274/8, aggrin_queue=7558127, aggrutil=99.85%
  sdb: ios=867359/3, merge=0/0, ticks=7558274/8, in_queue=7558127, util=99.85%

 

[Windows7ge]

1 minute ago, unijab said:

I avoid windows, so i dont use smb

 

I use targetcli to export "luns" from my storage system to my proxmox system.

 

I have sync=always and a ssd drive for slog.

 

I max'd out ram on my motherboard so I dont have any l2arc configured.

 

heres a recent fio test from within a VM with a lun passed thru.

I also have a PROXMOX server and I did attached a L2ARC/SLOG device to the primary pool but not for the application of a network share.

 

Admittedly I had not tested NFS. I tried enabling it on Windows within it's features menu but it was utterly broken.

[Bramimond]

24 minutes ago, RastaGrzywa said:

Linus, maybe You should consider tape drives for long term (archive) storage system?

Tapes are only good for backups. If your data isn't conveniently accessible it might as well not exist.

[maxhan]

Why do they not compress their project in a .zip/.rar-File when they are done? Would save some space...

[RejZoR]

31 minutes ago, Miracle Nachos said:

I'm a hoarder, but I still freak out at buying a HDD bigger than 2TB.

I feel like those guys at the birth of the Steam Engine who were terrified that going faster than 10 miles an hour would cause your intestines to shoot out of your anus.

Don't worry, just buy whatever you want, even a 16TB Exos drive. Just make sure to run Crystal Disk Info (CDI) as resident app so it constantly monitors HDD vital signs and temperature. In case something weird is starting to happen with the drive, CDI will notify you about it. The thing is, it's really rare for mechanical HDD to just die out of the blue. It'll always start doing something weird like funny noises, slow down unreasonably, have very long startup (spinup) times, click weirdly or constantly do error corrections. Combining warnings that you can hear and SMART info from CDI and you can spot a funny operating HDD and replace it before it fails catastrophically and you lose the data. The cost of new HDD is nothing even if you proactively replace it on first signs of problems over running two drives in parallel as redundancy. Now, if you have super critical data you still do that. But for home use, CDI approach is imo good enough.

[sub68]

ok wow one petabyte not enough go triple 

[G4te_Keep3r]

17 minutes ago, maxhan said:

Why do they not compress their project in a .zip/.rar-File when they are done? Would save some space...

With video you wont see much space saved if any. On my systems I do search for text iso vdisk files and compress them as they compress well. Have not analyzed a premier project folder but the amount of non video space is probably negligible to the point of not being worth it at all.

[Boomer0108]

Question:

I've always heard "RAID ISNT A BACKUP!"

So, how do you backup a beast like this?

[BachChain]

1 hour ago, Miracle Nachos said:

Here's something I never thought of before: Do you think we'll ever see 3.5 Inch SSDs?

They already exist, and they're terrible.

[G4te_Keep3r]

My Unraid server is basically at limits again (poor cache drives). Down to 1 drive bay free in my r720XD (not shown is the unassigned devices 500 GB SSD for VMs), and then there is the almost 30 TB on external drives (which includes an Unraid server (trial license there) of 250 GB drives...not powered on unless needed). Remember the iJustine tour, imagine that but mix of internal and external drives (and tmp folders borrowing storage on other peoples computers too (with permission FYI)).

 

My youtube has been in "I need to edit" mode for a few years as the multiple projects folders have moved between drives and new projects created and filmed. Well 1.5 vlogmas (1 and part of another that was cut short because it was a boring December) that were done all on my phone was uploaded, but Premiere is so much better than mobile editing and all the other problems with using just 1 phone. Right now most of the projects are together in a YT share on my server but random stuff like "gopro dump 05" and "goggles from event"..., so working on sorting/relabeling and hopefully to start editing (when not working on research project to graduate). At 6:30 in video, NAS building collabs were mentioned, so figured it couldn't hurt to post. Come down to Texas @LinusTech, have a Whataburger (being a Texas only fast food), and fix my server/setup before I finish my masters degree this summer.

 

Ignore the uptime, had a power outage just over 2 weeks ago and I still do not have a UPS yet.

[sub68]

wait each exos drive is 400 bucks

also each ssd is 1000 dollars 

[G4te_Keep3r]

43 minutes ago, Boomer0108 said:

Question:

I've always heard "RAID ISNT A BACKUP!"

So, how do you backup a beast like this?

"RAID (Redundant Array of Inexpensive Disks^[1] or Drives, or Redundant Array of Independent Disks)" (lazy so copied the definition from wiki)

RAID isnt backup, combining local drives into 1 storage volume and redundancy (speed depending on number of drives and config). Data drive fails, insert new data drive and rebuild what was there from the parity and other data drives (or something like that depending on config). True backup follows the 3-2-1 rule, 3 copies on 2 mediums with 1 offsite, reasonable large storage should be ok with raid redundancy and offsite copy. No idea what the industry rules are that google and all them use tho, but probably way out of most any consumers budget.

 

They talked about the "right" ways and the hack here.