bizaflare Posted July 15, 2019 Share Posted July 15, 2019 Hello everyone, after not having used my windows server machine with flexraid for 2 years, this weekend I decided to switch to unraid. Therefore, I cleared all its contents and put it on other harddrives in my network. Installed and configured unraid, installed S3 plugin, tried it out, created my array, created some shares. My setup is a small ASRock Q1900-ITX box, running five WDC WD30EZRX 3 GB HDDs and a (too small) 64 GBs cache SSD. 2 HDDs are configured for parity. Unraid Version 6.7.2. During creation of the array, disk 1 showed read errors for the first time and it showed unmountable. Thus, I ran a quick health check that passes. Tried it again and, voila, now it worked. Then I began transferring my data back to the server. First, the most important user data that I decided to put solely on drive 3. Besides some initial issues with the cache drive being full, no issues. Now I started to transfer back my media, which is supposed to go on drive 1 and 2. As I have all my data still backuped, I wanted to speed things up, by stopping parity calculations which was expected to last for days. I started to move data over - but after some time, read errors appeared on disk 1!? I now stopped the transfer, and did a quick and an extended SMART report check. Both of them passed (even though the extended one reads as there were errors found!?). Now I am puzzled and need advice what to do: Is disk1 trash and i should replace it? Just remove it and try with the remaining 4 (should i recreate my array then)? Try to shuffle disks (maybe use it for parity?) and try again? Smart report and analytics attached. Best regards dun-crimthainn-diagnostics-20190715-0442.zip dun-crimthainn-smart-20190715-0627.zip Link to comment
Frank1940 Posted July 15, 2019 Share Posted July 15, 2019 I had a quick look at this board. Apparently, it has -four- SATA ports. What are you using to provide the other two ports? Is disk1 connected to that device? Another question. What are the spec's on the power supply that you are using? It is recommended that Unraid server using power supplies that have a single 12V buss. There is a bit of slack in this requirement but, in that case, each buss must have the capacity to supply the current (including the starting current) of all the hard drives. Link to comment
JorgeB Posted July 15, 2019 Share Posted July 15, 2019 Despite the passed extended test, it looks like a problem with disk1, there were media errors (UNC @ LBA), you should keep an eye on it and replace if more errors. Sil controller is old and not really recommended, some models could cause data corruption when both disks are accessed simultaneously, if you can replace it. Link to comment
bizaflare Posted July 15, 2019 Author Share Posted July 15, 2019 5 hours ago, Frank1940 said: I had a quick look at this board. Apparently, it has -four- SATA ports. What are you using to provide the other two ports? Is disk1 connected to that device? Good point - I didn't remind having this issue when I built the system in 2014. I am using a Delock 95233 2 Port PCIe 2.0 Mini Card to run the other two ports. Apparently they are mapped to sdb and sdc, i.e. both parity drives. I wonder if this is a good idea or if I should diversify anyway, i.e. only one parity device attached to this device? 5 hours ago, Frank1940 said: Another question. What are the spec's on the power supply that you are using? It is recommended that Unraid server using power supplies that have a single 12V buss. There is a bit of slack in this requirement but, in that case, each buss must have the capacity to supply the current (including the starting current) of all the hard drives. I am not sure whether I fully get your question. I am using a 300 Watt 300 Watt be quiet! SU7-300W power supply from which I use its four S-ATA plugs (parity 1 and 2, cache SSD, disk 3) and an IDE to 2x S-ATA adapter (disk 1 and 2). 4 hours ago, johnnie.black said: Despite the passed extended test, it looks like a problem with disk1, there were media errors (UNC @ LBA), you should keep an eye on it and replace if more errors. Sil controller is old and not really recommended, some models could cause data corruption when both disks are accessed simultaneously, if you can replace it. Apparently, disk 1 is not behind the PCIe raid controller, though. This means that this shouldn't be the issue here!? As my mainboard does not have more slots I will have to use one if I want to stick to 6 drives. Do you advice in getting a better one? I have not yet transferred all my data. And having 6 TB of storage space would be enough for my current needs. Shouldn't I recreate the array, remove disk 1 and try without it as long as i have the chance? Link to comment
JorgeB Posted July 15, 2019 Share Posted July 15, 2019 4 minutes ago, bizaflare said: Apparently, disk 1 is not behind the PCIe raid controller, though. This means that this shouldn't be the issue here!? No, like mentioned it appears to be a disk problem. 5 minutes ago, bizaflare said: Do you advice in getting a better one? If two ports are enough get an Asmedia based controller. 6 minutes ago, bizaflare said: I have not yet transferred all my data. And having 6 TB of storage space would be enough for my current needs. Shouldn't I recreate the array, remove disk 1 and try without it as long as i have the chance? If you don't need the space then remove it, it's more likely to fail again. Link to comment
bizaflare Posted August 6, 2019 Author Share Posted August 6, 2019 Thank you for your help. The disk actually seemed to be the issue here - i removed it. No issues with read errors with the other disks so far. Link to comment
Recommended Posts
Archived
This topic is now archived and is closed to further replies.