What is your current system? How old were the drives that failed? What kind of drives were they? By robust, do you mean more highly available or just less failures?
What is your current system? How old were the drives that failed? What kind of drives were they? By robust, do you mean more highly available or just less failures?
Good Morning,
My current system is I7-6700T @ 2.8ghz (lenovo tiny chassis)
32 gb ram
256 gb m.2 sda
7 remaining drives all of various types and sizes in 2 Mediasonic Probox 4 Bay chassis'
2 WD10's 10 TB (newest drives place Oct 2021)
3 WD 8's "REDs" (shucked from externals) had 5, 2 failed production dates are all around aug 2017 i'm expecting the remaining to fail.
Hitachi 4tb (MJ0351YNG970TA)
HGST 4tb PL2331LAH3LZPJ
When I mean robust I primarily mean less failures or if a failure happens it's not to the point that I'm thrown into emergency mode because 1 drive can't / won't respond. If i can make it HA also cool but if it's going to increase the cost dramatically it's not worth it.
thinking of getting Seagate 16TB HDD Exos X16
3 WD 8's "REDs" (shucked from externals) had 5, 2 failed production dates are all around aug 2017 i'm expecting the remaining to fail.
I haven't had good luck with drives from usb enclosures either. And I usually try to replace (or at least move them to something less important) drives by the time they hit five years old. Do you shut your system down a lot?
When I mean robust I primarily mean less failures or if a failure happens it's not to the point that I'm thrown into emergency mode because 1 drive can't / won't respond.
Nothing OMV can do about drive failures other than watch smart to see if they are failing. And while it isn't supposed to work, adding nofail to the mergerfs options has worked for some people to avoid emergency mode.
if it's going to increase the cost dramatically it's not worth it.
HA always cost more since you have hardware just providing redundancy.
Seagate 16TB HDD Exos X16
I have an Exos X18 (for a couple months) and it good so far.
Would the raid with mergefsfolders be a bad idea?
If you are using raid then you don't need mergerfs folders unless you were going to have a bunch of raid 1 arrays. I would not use raid with those enclosures if they are connected via usb.
If you are using raid then you don't need mergerfs folders unless you were going to have a bunch of raid 1 arrays. I would not use raid with those enclosures if they are connected via usb.
Yeah they are usb enclosures currently as the tiny does not have esata connectors. I was thinking raid 6
I was thinking raid 6
Not with usb connected drives.
If you use a wildcard for your folders to merge (/srv/dev* or something), then I don't think it will ever fail to boot since nofail should be in the filesystem mount options already.
If you use a wildcard for your folders to merge (/srv/dev* or something), then I don't think it will ever fail to boot since nofail should be in the filesystem mount options already.
I will attempt that, Now to see if I can recreate by cron job to extract all downloads that haven't already been extracted. I used to have it but I have lost the code in the HD crash. I'm thinking something like
unrar e -r -o- /srv/driveident/downloads/ *.rarNot with usb connected drives.
Have you attempted Raid with esata or is that a bad idea also.
Have you attempted Raid with esata or is that a bad idea also.
I used it many years ago. It isn't my favorite thing but it is much better than usb.
Sie haben noch kein Benutzerkonto auf unserer Seite? Registrieren Sie sich kostenlos und nehmen Sie an unserer Community teil!