A community dedicated to the profession of IT Systems Administration
No generic Lemmy issue posts please! Posts about Lemmy belong in one of these communities:
[email protected]
[email protected]
[email protected]
[email protected]
I don't know, but ideally that data would be cached in RAM. Maybe if you used intelligent tiered storage with a flash tier it could reduce wear and access times.
Ultimately I doubt that this is going to have a significant impact on drive lifespans. A surveillance camera PVR is writing 24/7 which is more intense, and those drives still last plenty long.
Interestingly enough, there are HDDs purpose made for surveillance (eg. WD Purple), and their special feature is that they're dumb as bricks: since surveillance more or less continually writes, and only really reads when user directed, there's practically no start-stop-move head, no predictions, no sleep, no need to cache system files... Just write-write-write in a line, then when you run out of space, start over.
I don’t know, but ideally that data would be cached in RAM.
Not feesible, unfortunately, if we are talking about multiple terabytes of data.
Maybe if you used intelligent tiered storage with a flash tier it could reduce wear and access times.
Could you clarify what you mean?
A surveillance camera PVR is writing 24/7 which is more intense, and those drives still last plenty long.
That's a fair point; however, I have seen special hard drives exactly for this purpose.
I would go with ZFS via Truenas. It makes the setup pretty simple and it will have all the benefits of zfs
Torrents are never equally in demand. A large amount of ram could maybe cache the majority of reads, even to a multi-TB array.
I have seen HDDs advertised as WI (write intensive), RI (read intensive), and MU (mixed use). The advertising says that the WI drives will last longer under write intensive loads. I don't know how much truth there is to that.
That's for SSDs which is a different thing.
Hard drives are quite reliable these days. According to the Backblaze stats, the annualized failure rate for modern drives is only about 1.5%. And these guys beat the living shit out of their drives.
Thanks a bunch for that link! That's a really useful resource!
For an SSD: not really, in theory.
For an HDD: kinda. Spinning up and spinning down the disk technically always comes with the risk of the drive damaging because of the physical components involved, and will eventually wear out. Constant writes would definitely be far harder on it, but more spinning time is always generally likely to wear it out faster.
Spinning up and spinning down the disk technically always comes with the risk of the drive damaging because of the physical components involved
Ideally, the seebox would maintain a 100% uptime.
Constant writes would definitely be far harder on it
Would there be a difference for constant reads (reading is what the seedbox would primarily be doing)?
Constant reads wouldn't be as hard on the drive, but again, the more the mechanics inside the drive work/move, the more they will wear down. For HDDs, most failures are mechanical failures.
That said, even with a consumer grade drive, I personally wouldn't worry too much about it; modern drives are pretty solid in general, just make sure you backup anything important.
If you're really worried about it, WD's gold line is made for constant reads/writes 24/7 and to be reliable under those conditions
Use zfs with plenty of ram. The ram cache with help with speed and reliability
Just out of curiosity, why are you setting up a seedbox in a enterprise environment?
The ram cache with help with speed and reliability
Are you inferring that the torrents would be stored in ram? That would not be feesible with large amounts of data.
Just out of curiosity, why are you setting up a seedbox in a enterprise environment?
What do you mean? What enterprise environment?
This is a sysadmin sub
Ah, sorry about that. That's why I mentioned the 3rd paragraph in my post; I wasn't sure if this was the correct place for this post - I wasn't sure where else to go.
Yes, but the number of hours they can withstand these reads is rather insane. I've seen SAS level drives with millions of hours of runtime and no bad blocks. They are pretty robust these days!