busthead Posted January 23, 2019 Share Posted January 23, 2019 (edited) Hello, I moved the four drives in my existing array from built-in motherboard SATA to a Dell PERC H310 (taped and flashed to IT-Mode) in preparation of adding additional drives. Parity check progress was estimating weeks to months to complete. I let it run 8+ hours overnight and it was 29% complete (Average speed: 135 MB/s). I have shutdown the system. Can you please check the attached diagnostic and let me know if there is an issue that is causing the parity check to take so long? System Info: The H310 is a 8-lane PCIe v2.0 card. The motherboard supports PCIe v2.0. Therefore, the theoretical throughput (of the controller card) should be x8 (eight lane) v2.0 PCIe slot: 8 * 500 MB/s = 4000 MB/s = 4 GB/s = 32Gb/s. What is the expected throughput during a parity check with four 5,400 RPM SATA hard drives? Thx tower-diagnostics-20190123-0722.zip Edited January 23, 2019 by busthead Clarification and additional question Quote Link to comment
JorgeB Posted January 23, 2019 Share Posted January 23, 2019 Didn't you notice these? Jan 23 07:08:05 Tower kernel: md: disk1 read error, sector=4537757376 Jan 23 07:08:05 Tower kernel: md: disk1 read error, sector=4537757384 Jan 23 07:08:05 Tower kernel: md: disk1 read error, sector=4537757392 Jan 23 07:08:05 Tower kernel: md: disk1 read error, sector=4537757400 Jan 23 07:08:05 Tower kernel: md: disk1 read error, sector=4537757408 Jan 23 07:08:05 Tower kernel: md: disk1 read error, sector=4537757416 Jan 23 07:08:05 Tower kernel: md: disk1 read error, sector=4537757424 Jan 23 07:08:05 Tower kernel: md: disk1 read error, sector=4537757432 Jan 23 07:08:05 Tower kernel: md: disk1 read error, sector=4537757440 Quote Link to comment
busthead Posted January 23, 2019 Author Share Posted January 23, 2019 I had not, thank you. Will run a few tests on that drive a report back. If I replace that drive with another, of at least the same capacity, will the array rebuild itself? Thx Quote Link to comment
JorgeB Posted January 23, 2019 Share Posted January 23, 2019 Yes, it will rebuild the disk, disk has a high Raw_Read_Error_Rate, which is consistent with the issues you're seeing, like slow sectors and even bad ones, though since there are no read errors or pending sectors on SMART only an extended test would confirm, alternatively swap both cables with another disk and run another non correcting check, if errors persist it's the disk. Quote Link to comment
JorgeB Posted January 23, 2019 Share Posted January 23, 2019 Oh, and make sure you enable system notifications so you'll get a warning when there's a disk with read errors. Quote Link to comment
Recommended Posts
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.