CitroJim's AX, C3 Picasso, Cycling and Running Tales

[mickthemaverick]

[CitroJim]

With you having been the owner of both the "French" Cars in the picture, just look at how small the orange thing from Italy is in comparison to both the C1 and the C3 Picasso!

It makes the C1 look huge!!!

Safety is off..... :missile:

The great thing is that we all have our own likes, dislikes and opinions and are very free to express them; we all attend a very broad church

A cautionary tale of the propensity of a Solid State Drive (SSD) to suffer 'bit rot' if not kept in regular use has been very clearly observed today...

It's been ages since I've fired up my Diagbox laptop and thought I should, taking into account the above... Too late

It seemed to boot normally initially and I could start Diagbox in the VM. Then the underlying OS started to reboot at random short intervals, blue screen and occasionally claim the hard disk was missing. Finally, it became impossible to start the VM.

A Windows update resolved the issues with the underlying OS but the Diagbox VM was sufficiently corrupt it needed a reinstall from scratch; even the source files were corrupt and failed to extract in WinRAR with CRC errors. Good job I had the files safely backed up on an old-school hard disk!

All is good again but it does underline how SSDs should be kept in regular use and never used for archival storage...

I must now remember to fire up the Diagbox laptop regularly...

The SSD in use in the Diagbox laptop is identical to the ones fitted to both my constantly in use Win. 10 laptops and they have been fine.

[myglaren]

That is interesting Jim. Had heard of bitrot but hadn't connected it to SSDs.
I only have spinny disk machines and all backups go to spinny disks too.

[CitroJim]

That is interesting Jim. Had heard of bitrot but hadn't connected it to SSDs.
I only have spinny disk machines and all backups go to spinny disks too.

A well-known problem nowadays Steve, and with almost all new devices using them, one that's likely to grow. I was a little surprised at how quickly it happened.

Yes, definitely back up onto traditional hard disks. USB sticks and SD cards can also suffer bit rot so they're not good for backup either - except short-term. I've had SD cards rot and fail.

I back up to both a constantly running little Linux server and two flavours of the 'cloud' - Giggle and Microsoft OneDrive. I think I'd be hugely unlucky if they all failed...

I sourced my SSDs from a company that came with very good recommendations from the good folk of the vintage radio forum. Generally they know their stuff on there...

I've had one SSD fail completely but not one from that source; it was the one that originally came with my eBay-special laptop. It's also consumed a keyboard.

As Win 10 nears the end, I'll be treating myself to a brand-new Win 11 (maybe 12 by then!) laptop of good quality and relegating the best of my old Win 10 machines to Diagbox use.

[mickthemaverick]

I'm still a dinosaur where backups are concerned. I have 3 x 128 gig usb sticks with velcro labels G F S which move around each time I backup, usually fortnightly. For the non IT literate they stand for Grandfather, Father and Son as was always the way when I started computing.

I also use the oldest form of backup for my three ultra important data files, my passwords, my contacts and my calendar, I print them and keep the paper in my loft mounted wall safe. I generally only print them when I have had to make an important change to one of those files but I only reprint the page with the alteration!!

Yes I'm OTT and meticulous but that is what comes from my second career in security!!

[Stickyfinger]

Now we ALL know Mick......as does all the "internet"................ You will need to change everything now

[mickthemaverick]

To paraphrase Sergeant Schultz "You know nottin"

[Stickyfinger]

All I will say is.....get a Drone Scanner.....it will keep you safe...

[MattBLancs]

Enjoyed the change in direction from things in a wheelbarrow with a rough resemblance to gooseberries to Jim's IT saga

Spoiler: show

"oh, bollocks!" Twice!!

Not aware of SSD "bit rot" so I'm learning something new as usual

[myglaren]

I have three USB sticks that have died but no SD or similar cards. Oldest is an XD card in my Olympus DSLR, probably 20 years old. It has two cards in it, the XD and a CF. Normally only use the XD.

[Zelandeth]

Big question there Jim, is while they're badged as being from a specific manufacturer, where were they sourced from?

If Amazon or eBay I'd say there's at least a 1 in 4 chance of it being counterfeit these days and actually just the cheapest of brand less tat with fake badging stuck on.

I've had this with electronics parts, screwdriver bits and even a bloody bunion file in the last six months where what turned up was plainly AliExpress/Wish.com/Temu grade tat claiming to be from a name brand.

I know SSDs aren't suitable for archival storage, but even so it shouldn't die if left sitting for a couple of months. Something is far wrong there.

[CitroJim]

I know SSDs aren't suitable for archival storage, but even so it shouldn't die if left sitting for a couple of months. Something is far wrong there.

I'm in agreement with that Zel... It was about 3 months I'd say... And I think it's a bit poor.

These are the disks in question:

I bought three in total, all from the same eBay supplier. The ones heartily endorsed by a number of the Vintage Radio forum members and from their recommended supplier, being aware of the fakery that goes on...

Who knows, the one in question might be a fake... A good chance I suppose... They were not massively expensive.

There was a very long, detailed and interesting thread on there about SSDs, their fakers and susceptibility to bit-rot...

I can't post the link here as it's in the members-only area and one only accessible after a minimum number of posts have been made...

Two of mine were 512GB ones, bought a few days apart, and have been, touch wood, faultless in my regularly used Win 10 Laptops and this one pictured, a 256GB one, was bought a bit later for my Diagbox machine...

From now on I'll fire up my Diagbox machine on a regular basis and see how it goes... Given the behaviour of the Diagbox machine, my first thought was an intermittent hardware fault but after updating the underlying OS it then ran fine for over 24 hours so that's now less likely. Only time will tell on this one...

[Zelandeth]

Interesting. Only ones I've bought in the last 10 years have been Samsung or Western Digital. Thus far no issues to report.

One has a relatively low state of health reported through the self monitoring system. Though I reckon that's largely down to having spent three or four years 24/7 in a system that was too old to have an SSD aware BIOS, so the Trim functionality and other OS optimisations to help look after an SSD weren't available. There were a number of cludges involved in that machine - the oddball 32-bit UEFI boot environment (it was a 64-bit machine) being one of the most annoying as it requires non-standard install media to make things actually boot. I guess that's what you get with old Apple hardware though when you're not running their OS. Credit where it's due though - I was still using it as my main machine until less than two years ago. It also cost me nothing saved for the £30 or so of ECC memory I needed to max it out at 64Gb. Was purely the CPUs that really were starting to be a bottleneck though. A pair of Xeon 5365s were stormers in their day (4 cores each at 3GHz), their architecture was just showing it's age. Oh, and it's ability to turn the computer into a very effective space heater when busy. Still a lovely bit of industrial design though.

I can't actually remember which one that is...I *think* it's the one which currently has Windows on. I'll need to have another look at the Smart data to see if the state of health reading has changed since I last looked. I'm guessing the answer is most likely going to be a no though.

Of course unlike Windows, at the very least all the mainstream Debian based Linux distributions do read that data, and will pop up warnings if anything untoward is recorded. Something which did save me a bunch of hassle ~15 years ago when I had a hard drive randomly die. It noted the throughput performance had dropped sharply, warned me that the drive was likely failing, and gave me time to take an updated backup before it died completely (or rather was wiped then sent back to the manufacturer as it was still under warranty - by about a fortnight!). I really don't understand what possible excuse Microsoft can justify as their reason for not giving you access to that diagnostic data. I guess it's only a 30-ish year old protocol, maybe still too new fangled for them.

[CitroJim]

No error messages were seen about the SSD Zel but now you have me pondering a thought: the SSD in question is in a very old HP EliteBook I use to run Diagbox and I'm wondering if the BIOS is SSD unaware and this may be the root cause of the problem?

Touch-wood I've not had a proper mechanical hard disk fail for years and I do wonder if that's because they've ben in machines running 24/7?

Yesterday I reset Bluebell's clock back to GMT by removing the battery. Whilst the battery was out I took the opportunity to apply a bit more rust preventative in the area.

A lot of yesterday was spent making the most of the gorgeous late autumn weather in the garden, at parkrun and out on the bike

[Zelandeth]

No error messages were seen about the SSD Zel but now you have me pondering a thought: the SSD in question is in a very old HP EliteBook I use to run Diagbox and I'm wondering if the BIOS is SSD unaware and this may be the root cause of the problem?

Touch-wood I've not had a proper mechanical hard disk fail for years and I do wonder if that's because they've ben in machines running 24/7?

Yesterday I reset Bluebell's clock back to GMT by removing the battery. Whilst the battery was out I took the opportunity to apply a bit more rust preventative in the area.

A lot of yesterday was spent making the most of the gorgeous late autumn weather in the garden, at parkrun and out on the bike

If it's a Windows machine, unless you're running a third party tool to watch the SMART data, or the system BIOS is clever enough to warn you at startup that something is amiss you wouldn't have had any warnings delivered.

I just had a look at the drive I'd mentioned above, and the values look basically identical to when this current machine was put together and that drive was moved over to it about a year and a half ago.

I wouldn't expect it being a non-SSD aware machine to be a big issue for something that's only seeing occasional use. That side of things is more likely to be a problem for machines that are in continual use as things like continual writing of system logs rather than batching things, background file system defragmentation, hammering of the page file (yes Windows, I'm looking at you...WHY were you even USING virtual memory when the machine had 64Gb of physical RAM?) continually are things which will wear out an SSD prematurely as there are a finite number of times that each of the stored bits can be flipped. For something that's only being used for a couple of hours here and there that really shouldn't be an issue.