[Tim Jones]

Thanks, @Jim_Curtis. Also, SoftRAID 6 is nearing release and new RAID 5 support is coming with the update along with Big Sur (except 11.2 – kernel bug – being fixed in 11.3) and M1 support. To clarify, if you’re using SoftRAID under Big Sur – please hold off on the 11.2 update and wait for 11.3 to release.

@Bob_Zelin – RAID 0 is very effective when performance and capacity are more important than redundancy – especially if you also have a backup plan in place to another environment (like LTO 🙂 ).

[Tim Jones]

Of course, my initial recommendation would be to look into an LTO tape solution. The initial cost of entry may be a bit frightening, but the long term costs and peace of mind will pay off quickly.

For 40TB of data, you could use 7, 6TB LTO-7 tapes that are –

• Easily transportable
• Robust even in India’s summers
• Require no power to maintain
• Can be stored safely offsite in a vault or even duplicated to provide a geographically distant deep archive.

And with appropriate connectivity (TB 3 or SAS) at home and your studio, you can carry the tape device between the systems so you only need a single tape device.

Aside from that, you don’t mention the network capabilities between your home and studio. If you have a high speed service, you could take the disk route and place a low-cost disk array in both locations and then use something like rsync to keep both side sync’d so that you always have the same data available in both locations.

In either case, you’ve provide at least a minimum of geographical (and hopefully power grid) separation between your data while making it easily available at both ends.

Good luck!
Tim J

[Tim Jones]

No Re-read necessary. Your post was obviously intended to spread FUD and is nothing short of misleading in nature. Rather than turn this into an even grander argument thread, I’ll state simple facts with regard to the meat of this thread:

OWC is fully prepared to support not only a standalone, Thunderbolt 3/4 tape device, but also the newer tape technologies that we currently have in development. As with any major company’s efforts, missteps such as Todd’s communication with our support team can occur, but it’s what we do to correct and resolve those missteps that matter most in the long term.

Regarding TOLIS Group’s demise and the rumors of BRU’s death, BRU is not dead nor does it need to be resurrected. The same design and coding team are also part of OWC and continuing the development and growth of the BRU software solutions.

My summary point to you, @timgerhard, is that we can all coexist here on the forums, as long as we don’t try to use negative representation and FUD to defend a point. Stick to answering generalized questions about the technologies and specific questions that pertain to your products and efforts, and we’ll all be fine here.

—

Tim Jones

Chief Solutions Architect
Other World Computing, Inc.

https://owc.com, https://macsales.com

[Tim Jones]

you can see that this product comes with myLTO software from Imagine Products. And OWC is also recommending Pomfort Silverstack XT, Hedge Canister, YoYotta LTFS, XenData6 Workstation, Archiware P5, and more.

So I think it’s pretty clear as to the direction that OWC wants to go with this product line.

As for what it includes, the myLTO software is an option that we happen to sell, it’s not included by default with the purchase. BRU solutions will also be an option, so we’re happy with whatever the customer needs to use in their environment.

—

Tim Jones

Chief Solutions Architect
Other World Computing, Inc.

https://owc.com, https://macsales.com

[Tim Jones]

Thunderbolt 3 devices require a strict certification for both Windows and MAC. OWC has not passed certification for Windows, so that’s why you’re probably not getting anywhere. MagStor is the only Thunderbolt 3 drive certified for Windows, and also has Apple M1 compatibility to boot.

Absolutely not correct Tim and you need to retract that statement and the implication that it creates. Our lack of a Mac M1 driver has no bearing on our certification on Windows, it has to do with a bug that we’ve discovered in the M1 STUC layer. One that will probably affect your driver as well, since it’s related to the kernel.

Next, BRU is NOT dead software. it is now owned by OWC, I am still it’s primary architect, and we are continuing in its development and growth based on its 36 year legacy of reliable backups. Also, BRU solutions have no dependency on open source software that can break depending on the version used or dependent packages like ICU and FuSE. You should really use caution when making such statements.

Your performance statement is also FUD and needs to be clarified. You are correct that the half height drives are only rated for 300MB/sec native throughput like its LTO-7 older brother and we will update that on the website, but with BRU driving the I/O instead of LTFS, we’re easily achieving 325MB/sec+ processing average M&E file data with the half height LTO-8 drives.

Finally, OWC’s product line breadth has no bearing on how we support a new product. Again, your deliberate use of misleading statements needs to be retracted.

—

Tim Jones

Chief Solutions Architect
Other World Computing, Inc.

https://owc.com, https://macsales.com

[Tim Jones]

@Terry – if you would like to give BRU PE (now ArGest Backup) a try, the demo is still available from the TOLIS Group website:

https://www.tolisgroup.com/requestdownloads.html

OWC will be fully supporting the BRU solutions and continuing the 36 year development heritage.

Tim

—

Tim Jones

Chief Solutions Architect
Other World Computing, Inc.

https://owc.com, https://macsales.com

[Tim Jones]

Hi Todd,

We are working on the driver and have run into an Apple kernel problem that we are working with Apple to sort out. Hang in there as the solution is coming.

Also, the BRU PE update to ArGest Backup works properly under Big Sur and is a Universal app for the M1 units.

Tim

—

Tim Jones

Chief Solutions Architect
Other World Computing, Inc.

https://owc.com, https://macsales.com

[Tim Jones]

Another main issue that I am facing is while I am trying to backup data, for 2.31gb file it is occupying the space on tape is 4.4gb. I have given the remote control to the provider and he finally comes to an conclusion that it is the issue with the tape media. But I have even tried with the old tapes as well which already partially backed up also resulting the same. Couldn’t find a solution.

In light of the numbers above, when you “overwrite” an LTFS volume, you’re losing the capacity of the data that you’re overwriting because, unlike a disc, LTFS doesn’t replace files that were already written, it just marks them as not existing in the LTFS index, resulting in the loss of the originally used space on the tape. If you tried writing that file to the tape again, you would discovered that you have lost another 2.31GB.

As for formatting, I go back to my original recommendation. In a Terminal (not the HPE UI), execute that command with the tape inserted in the drive but NOT mounted:

mkltfs --wipe

That will properly remove the LTFS formatting and effectively “erase” that tape. You would then reformat that tape to start fresh.

As for the other issues, what is your environment like – temperature, humidity, dust? All of these things can negatively affect the operation of a tape drive.

Tim
—

Tim Jones

Chief Solutions Architect
Other World Computing, Inc.

[Tim Jones]

There are two erase modes for tapes – a quick erase and a full erase. The full erase (what you’re probably running) will take the full time to erase the tape that it takes to fill the tape with data. For an LTO-6 tape, this is a bit over 4 hours, LTO-7 is 6+ hours and LTO-8 is up to 10 hours.

On the other hand, the short erase only clobbers the beginning of the tape and the ability to retrieve data from the rest of the tape by creating a new EOD (end of data) marker in the first block.

LTFS tapes take this one step further since you either need to know how to change partitions before running the erase steps, or use a tool that allows you to delete the LTFS partitioning (effectively performing a short erase) and then recreate the LTFS format.

This last method is probably your best option since you’re using LTFS. With the HPE LTFS tools, use the mkltfs command:

mkltfs -w

or

mkltfs –wipe

And then reformat for you normal use.

[Tim Jones]

Terry – you’re probably witnessing the delays caused by the input pipe of the LTO-6 running dry (underrrun) due to the QNAP coming over 1GbE (100+/- MB/sec). To properly stream an LTO-6, you need to be able to supply a sustained 200MB/sec minimum. Otherwise, the drive has to pause, wait for the input buffer to hit its high-water mark, reposition the tape, and then continue. Combine that with the fact that you can’t adjust the I/O buffer sizes with LTFS-based solutions and the result is a much, much overall slower task.

With FCP projects containing lost of little files, this becomes even more apparent as it takes even more time to open, stat, read, and close each of those smaller files – further reducing the rate of data feeding the tape drive’s input buffers.

LTO-7 and LTO-8 are worse since they need over 450MB/sec sustained input to properly stream.

There really isn’t a way to speed things up unless you can bump your QNAP connection to 10GbE.