Running a windows 7 firmwre updater on RaspiOS

I think Seagate are specifically referring to drives currently on the market, not past models. It's also not unusual for manufacturers not to disclose about USB drives, only SATA drives. Previously SMR was only on very large drives (8+GB) but recently it's started creeping into smaller drives (to save costs, not that drive prices have come down much).

If you have the drive plugged in, run:

smartctl -a /dev/sdX

(may have to install smartmontools package) That should tell you the model number of the drive mechanism inside the USB enclosure. Then Google for that.

(it is possible smartctl may need an additional flag if your USB enclosure needs special handling to pass SMART commands - see the help for some examples)

Theo

Reply to
Theo
Loading thread data ...

Hard drive prices will probably increase as volumes decrease due to SSDs. They will only make the larger and more expensive spinning rust drives, and anything marketed in smaller sizes is probably going to be a larger drive where one of the platters has failed QA.

---druck

Reply to
druck

That doesn't help much. The drive model is ST1000LM048-2E7172 The two Seagate documents still disagree. It's possible TRIM support was added pre-emptively, since firmware is free once written.

Thanks for writing,

bob prohaska

Reply to
bob prohaska

AFAIK, SMR doesn't hurt performance until the drive starts re-writing previously-used sectors, rather like flash. If your drives haven't reached that point, they'll function roughly like CMR (also called PMR) devices. Given the capacity of SMR drives, it might make sense to just use them as approximately write-once devices.

Still, I'd like to figure out how to enable TRIM, which brings me back to the issue of running the enclosure firmware updater, or determining that TRIM is already enabled on the enclosure. It seems clear that the disk and RaspiOS both support TRIM, it's just a matter of turning it on.

Thanks for reading,

bob prohaska

Reply to
bob prohaska

It shouldn't be in drives intended for NAS usage. The 8-TB Toshiba N300s I bought a couple months ago don't do SMR, and it looks like they offer capacities up to at least 14 TB. Look for "CMR" as the antonym of SMR.

Looking at Toshiba's other drives, it appears the only ones that use SMR are their 2.5" drives. Seagate only uses it in 2.5" drives and their Barracuda and Archive drive families. Western Digital uses it in at least some 2.5" drives and some Blue and Red drives (what's troubling is that the Red drives are supposedly for NAS use, though all Red Pro drives and Red drives 8 TB and larger are CMR).

It would seem that avoiding SMR isn't as difficult as it may have first appeared, though you are going to pay a bit more for drives that don't use it.

_/_ / v \ Scott Alfter (remove the obvious to send mail) (IIGS(

formatting link
Top-posting! \_^_/ >What's the most annoying thing on Usenet?

Reply to
Scott Alfter

I've no argument with that reasoning.

But I suspect there are situations where SMR track rewrites have a greater throughput impact than that due to block remapping on SSD storage, simply because, while the way SSD drives remap blocks tends to distribute rewrites across the whole volume, if I've understood the process correctly, SMR rewrites tend to stay localised on or near the physical tracks occupied by the block being rewritten, so if a number of rewrites are done to physically adjacent data blocks, I can see a situation developing where te resulting series of track rewrites affect each other adversely, which could set off an avalanche of forced rewrites that really hammer throughput, not helped at all by the buffer tracks not being available to deal with unrelated writes and/or rewrites.

Or something like that. As a result, if you've got an EXT4 or other journalling filesystem on the drive, particularly if the partitions are relatively full, you'll get the performance slowdowns a lot sooner, simply because the rate of rewrites will rise much faster than it ever does for cold data (transaction logs etc) being recorded in a non-journalling file system where rewrites hardly ever happen.

Its somewhat unclear to me what exactly TRIM is meant to improve when used on an SMR drive: I can see that it would be a big help if it could defragment files etc, but surely that can't be done at the drive level since it implies some knowledge of the filing system's logical structure?

--
--   
Martin    | martin at 
 Click to see the full signature
Reply to
Martin Gregorie

It remains fairly difficult if I want a 2.5" drive rather than 3.5". Is there some special trick for finding them? By pure luck I found this:

formatting link

which reports that these drives use CMR: HDWJ110UZSVA, HDWK105UZSVA, HDWJ105UZSVA

Only the first is even listed on Amazon, with no sales nor reviews. It's particularly unhelpful that Amazon pads search results with unrelated products. An exact match search would help a lot.

There are many online lists of SMR drives, it's not so easy to find non-SMR drives, especially in smaller sizes, for sale.

If anybody has search terms or tips that help please post them. I'd like to stick with Amazon as the supplier.

Thanks for reading,

bob prohaska

Reply to
bob prohaska

l200/

From the stuff you posted it appears that all WD 2.5" 500GB Blue and Black drives and 1GB Red drives are CMR.

Similarly all WD 3.5" drives of 1TB or smaller are CMR and so are larger Red Pro, Black and Purple.

That is for drives made in the first part of 2020 - after that its anybody's guess

--
--   
Martin    | martin at 
 Click to see the full signature
Reply to
Martin Gregorie

Indeed, there do exist refurbs and old stock at 500 GB. Some of the new old stock is priced extravagantly, with neither reviews nor sales. No current, mainline products with clear specification that I can find.

It does appear that the best way to avoid SMR is to opt for 3.5" drives. Both selection and documentation are much better. It rather blows my "low power" goal out of the water, but I'm a niche buyer.

The trend to SMR is rather clear. I'd like to learn how to live with it, on the assumption that it's unavoidable for small, low power disks. It seems likely that documentation will improve as more people care. Perhaps SMR and CMR will become checkboxes on Amazon's product search 8-)

Most of my concern over SMR was related to use as a swap device. With the advent of the 8GB Pi4 the swap issue seems to be gone. If TRIM can be made to work with an SMR mechanical disk there might not be a problem at all.

Thanks to all for reading and replying!

bob prohaska

Reply to
bob prohaska

Googling indicates it's SMR:

formatting link

Theo

Reply to
Theo

On SMR, writing is expensive because you have to rewrite the whole track (and overlapping ones I think). TRIM means the drive has a view of what data is valid and what isn't. That means the drive won't have to rewrite data that isn't valid. It can aggregate writes into empty tracks so that it minimises the amount of rewriting it has to do - if the drive is 'empty' it doesn't need to bother at all, just fill up the empty space.

Like SSDs, it also means a more dynamic mapping of block addresses to physical location - rather than classical cylinder/head/sector addressing where there was a quasi-static mapping from block to location (unless you had sectors with errors), here that mapping gets permuted on a much more regular basis.

That's for Drive-Managed SMR. For Host-Managed SMR the filesystem does this, being better placed to optimise the process.

Theo

Reply to
Theo

Yes, and the product data sheet claims it's perpendicular:

formatting link

One of the criticisms directed at manufacturers in general and WD in particular was a silent switch to SMR. After complaints, docs were updated to reflect the change.

The gist is that published specs aren't necessarily a reliable guide. Eventually it'll likely get clarified, but for now, in the 2.5" 1 TB class of drives, SMR is difficult to reliably avoid.

Thanks for reading,

bob prohaska

Reply to
bob prohaska

At that size either you use an SSD (NVMe for preference) or you're penny pinching.

--
Steve O'Hara-Smith                          |   Directable Mirror Arrays 
C:\>WIN                                     | A better way to focus the sun 
 Click to see the full signature
Reply to
Ahem A Rivet's Shot

Price difference between HDD and SSD are fairly close. eBuyer is flogging:

--
--   
Martin    | martin at 
 Click to see the full signature
Reply to
Martin Gregorie

The latter.

8-)

bob prohaska

Reply to
bob prohaska

Yes. I've run this desktop now on SSD for as long as a spinning rust disk used to run and its SMART errors are well in hand.

Unless you want super high capacities, I see no point in spinning rust at all. I've 6GB of rust in my server, yes, but I would be happy to replace it with SSD when it starts to fail, if the price were similar.

--
Labour - a bunch of rich people convincing poor people to vote for rich  
people by telling poor people that "other" rich people are the reason  
 Click to see the full signature
Reply to
The Natural Philosopher

It's tempting to guess all three are old stock and won't be available for long. Perhaps they'll become expensive specialty items, or simply go away. In any case I'd like to find a storage solution with a few features:

Adequate capacity, say quarter to one TB Easy replacement availability Long(ish) lifespan, several years at 24/7 operation Power consumption at idle comparable to an idle Pi4 Performance sufficient to keep up with the Pi4 The latter two might be the most direct conflict.

As evolution plays out it may well favor SSD. For now HDDs seem to be developing the same handicaps as SSDs. I think HDDs still win on write cycles and power consumption. Not sure how they compare for practical life span. Life span is a function of overprovisioning for SSD, if SSD costs keep dropping they might win on both counts. Last I checked SSDs were a little worse for power consumption, but that was months ago and I don't know where the trend is going.

It's not clear where this goose chase will end... Perhaps at SSD. Perhaps at 3.5 inch HDD.

Thanks for reading,

bob prohaska

Reply to
bob prohaska

That struck me as odd, so I googled.

The balance of opinion is exactly the reverse. SSD uses up to 5 times

*less* power?

This spec

formatting link
ssd-2879-800092.pdf?_ga=2.57708201.28239759.1543588356-1527752628.1543588356

gives operational power on a SSD of ~3-4 watts and idle power 56mw. In most applications due to the high read and write speed the operational times will be very short.

formatting link

shows far higher figures for read write and idle...

--
The New Left are the people they warned you about.
Reply to
The Natural Philosopher

+1 exactly my thoughts. I am thinking of replacing the disks for the build environment 2x1T with SSD - there it would really make sense, but price is still too high even if one calculates the power consumption delta.
Reply to
Deloptes

formatting link

ssd-2879-800092.pdf?_ga=2.57708201.28239759.1543588356-1527752628.1543588356

formatting link

Absolutely correct and at present all servers I have seen use SSD 4 or 6 even some 8TB disks (like backup servers for example). I did check - those disks/models are not available on the consumer market (unfortunately).

At the end new technologies will win the game and in couple of years we'll be talking the same, but may be about SSD and NVMe or NVMe or whatever comes next.

As of my experience and a long research some years ago I use at home only WDC WD10EFRX-68xxxxx and WDC WD20EFRX-68xxxxx. The proved to be really good, I would say excellent. Some years ago when thinking of replacing some older disk I did a research on the 3, 4 and 6TB versions of same model - only negative experience and issues, so I preferred to stick to the two mentioned above.

I find this discussion extremely interesting. Thank you all

Reply to
Deloptes

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.