Admirably blunt!
Yes.
The processor is suitable (Windows processor requirements Windows 11 supported Intel processors | Microsoft Learn); there is a good reason to do the upgrade (impending loss of support) and there’s no reason no to.
Admirably blunt!
Yes.
The processor is suitable (Windows processor requirements Windows 11 supported Intel processors | Microsoft Learn); there is a good reason to do the upgrade (impending loss of support) and there’s no reason no to.
@billybutcher , @hairbear and @Porridge
That’s better - it’s what we wanted to know in the first place.
@billybutcher - good idea about a new system disk.
So now it’s just finding the time (in between putting in a new front and back door, repairing a large barn roof, finishing building a catio and preparing the house/garden for winter) .
Watch this space (in a few weeks) for how it went.
Thank you everyone. We really appreciate the comments, help and advice.
Apologies - it is a well-worn SF tradition for threads to drift off-topic (or at least into a slightly different direction…)
It’s also not really RAID because there’s no redundancy.
That is why RAID5 is out of favour for larger disks because data corruption becomes an almost certainty when copying multiple terabytes and disk failure under the much heavier than usual access is significantly more likely - as you experienced.
One issue is that people tend to buy disks to construct RAID arrays at the same time -which often means getting disks that were manufactured in the same batch - which increases the chance there will be multiple failures at the same time.
If you must have a parity array with drives >= 1TB approx then use RAID6 which can survive two disk failures.
Personally I have a rule of never using the same disk type/batch when building an array - so if it’s a mirror each drive will be a different manufacturer, if three drives in an array likewise - more than three and it gets difficult to have each be different but I still aim for a mix.
And, on command, one of my RAID disks has just failed and Amazon is kindly sending me a replacement by tomorrow. They tried to fob me off to Seagate initially and I had to remind them that, as the seller, they had first responsibility.
@ChrisMann - some of these comments made no sense to me whatsoever - that’s what you get for being a grey headed version of a dumb blonde - but my husband found them interesting !
Don’t worry - as I mentioned above, we were getting excessively geeky (and off-topic) talking about RAID hard drive setups!
For those still not clear about RAID levels (and should anyone have the slightest interest).
RAID0 - glue n same-sized disks together - any one dies and the array is toast.
RAID1 - duplicate data across 2 or more drives. As long as one survives data is safe.
RAID5 - At least 3 drives, extra “parity” blocks are stored which means any one drive can fail and data is safe
RAID 6 - At least 4 drives, two parity blocks so up to two drives can fail.
RAID 10 - combination of RAID 1 + 0. Number of tolerated failures varies depending on topology
RAID0 is only ever for short term “I really need a bigger disk for this temporary storage but only have two (or more) smaller ones”. Actually *increases* the chance you will lose data due to a drive failure.
RAID1 - as well as protection from failure potentially any drive can be taken away and is a complete copy of the data (trivial with Linux RAID but propriatiry solutions might not allow this).
RAID1 slows write speed as data needs to be copied to every disk in the array but read speed can be improved, especially on hard drives, as reads can be done in parallel across the array.
RAID5 - write is very slow as existing data needs to be read and combined with data to be written to calculate the parity block, then the new data *and* the updated parity block need to be written. Read speed normally improved compared with a single drive.
RAID6 - same as RAID 5 but two parity blocks so write speed impacted more.
RAID 10 - Take n RAID mirrors, then stack as a RAID0 array. Tolerance of disk failure varies - e.g with 6 drives making 3 2-drive RAID 1 arrays failure of a single drive will be tolerated, failure of the “wrong” two drives will lose data but failure of the “right” three drives wouldn’t.
The other issue is storage efficiency - RAID5 yields the largest usable space as only one drive is used for parity (it’s not a specific drive, parity blocks are spread across all drives in the array) - so three drives in a RAID5 array will yield 66% of the total storage as useable space. Space efficiency rises as more disks are added.
Mirrors (RAID 1 or 10) at most allow 50% of the total space to be used and the more mirrors in the array the less space efficient it becomes.
RAID6 with the minimum number of drives (4) is also 50% space efficient but improves as the number of disks increases.
Mirroring is probably easiest and can be set up with just two drives. I’d avoid RAID5 unless you want absolutely maximum space and/or only have three drives. If you have 4+ drives RAID6 is best for redundancy.
There are other RAID numbers - for the most part they are not used/interesting (well, sometimes RAID60 is used which is multiple RAID6 arrays themselves built up into a RAID0 array - see also RAID50).
Many, probably not. Personally, I do, at least on my Macs, and Linux workstations, even at home.