August 1, 20232 yr Hi, noob question about BTRFS here. I have a BTRFS cache pool in Raid 1, using 2 Crucial MX500s bought and deployed at the same time. I think one of them, the first disk, is about to die due to wear. My question is, would I expect the wear on the 2nd SSD to be similar? or would it be lower? I know that Raid 1 has the data mirrored across both disks, but I'm not familiar with how the mirroring works. I noticed that the 'writes' count is not the same between both drives in the pool. Thanks. OS: Unraid 6.10.3, if that's relevant. Edited August 1, 20232 yr by artkingjw typo
August 1, 20232 yr 30 minutes ago, artkingjw said: My question is, would I expect the wear on the 2nd SSD to be similar? It should, can you post diags to confirm the pool is only raid1?
August 2, 20232 yr Author Thanks. To add some detail, it is the 2nd cache pool, cache_n, i'm worried about. SMART data suggests that they have 23% life remaining. The first drive has had 2 BTRFS issues in the past 3 months. In the first incident, the drive was still detected, but I couldn't do anything the the data on it until I did a scrub+repair. In the 2nd incident, the drive became undetected, but unraid didn't seem to warn me? I swapped the SATA cables and ports around and it was detected again. I then did a scrub+repair. I haven't been a good admin, and allowed my array disks to get full, so all of my new data has been on the RAID1 SSD recently. I don't actively rewrite any old data, so my SSD cache mostly fluctuates between 60-90% usage as i work on my newer files. As I understand it, it's bad to keep writing to SSD's that are nearly full. I'll organize my data to free up space when I get more time soon... homecloudv2-diagnostics-20230802-1347.zip Edited August 2, 20232 yr by artkingjw detail
August 2, 20232 yr Pool is correctly configured for raid1 and both SSDs show the same wear, they both have 23% predicted used life, so very far from 100%, also note that reaching 100% doesn't mean it's going to fail, I've have one from the same model currently at 156% and still going strong. Since they are MX500 check if they need the firmware updated, older firmware versions have been reported to cause devices to drop, also check this for better pool monitoring.
August 2, 20232 yr Author 2 hours ago, JorgeB said: Pool is correctly configured for raid1 and both SSDs show the same wear, they both have 23% predicted used life, so very far from 100%, also note that reaching 100% doesn't mean it's going to fail, I've have one from the same model currently at 156% and still going strong. Since they are MX500 check if they need the firmware updated, older firmware versions have been reported to cause devices to drop, also check this for better pool monitoring. Oh? I thought 23% is what remains? That's what the name of the field is? "202 Percent lifetime remain". I am aware that, whatever it is, these metrics are just based on the 'rated' life, and in reality, many drives will exceed this rating, although that would be a bit of a lottery. Thanks for the tips.
August 2, 20232 yr 38 minutes ago, artkingjw said: I thought 23% is what remains? For this device it's means used, remain is the normalized value, this is form a device just recently added to my server, still at 100% remaining: And this is for the device that's well past 100% Note that the normalized value changed to 200, I guess to not show a failed SMART attribute, since threshold is 1
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.