On 18/12/18 22:04, Alexis BRENON wrote: > Hello everyone, > > TL;DR: > My 4 drives raid-5 array won't assemble on Debian. > Look at the commands outputs: > https://gist.github.com/AlexisBRENON/514500b90a477833c43584d62371925c > > --- > > It make some times (about 6 years) since I created my RAID5 array. > Recently, I started to get some strange behaviors with my HDD: UEFI > boot entry repeatidly erased, RAID component removed. In an attempt to > fix these, I upgraded my BIOS firmware (why not, it was very > outdated). So this computer is six years old? > > However, this did not fix anything. In another attempt to understand > what is happening, I run a SMART full test (from ArchLinux live USB > key) on all my HDDs. Only one of the four RAID drives reports a read > error (need to be replaced ?) as reported in this smartctl sum-up: > https://gist.github.com/AlexisBRENON/514500b90a477833c43584d62371925c#file-01-smartctl-txt > Read errors aren't necessarily a problem - it's if they can't be rewritten that it's dangerous. But obviously, if the data can't be read, it won't be automatically rewritten, you need to intervene manually ... > When booting on Debian, dmesg reported a ATA bus error. After some > investigations, this comes from a faulty SATA cable. Changing cable > fix it. > > Now, I'm here, with an up-to-date BIOS firmware and a 4 drives RAID-5 > with 1 drive reporting some SMART read error (and one drive aside for > the system (/boot, /usr, /var, etc.)). > > When booting debian, systemd timeouts failing to start some RAID > devices and thus not mounting the MD device. Below are the links to > outputs of most of the commands listed in > https://raid.wiki.kernel.org/index.php/Asking_for_help. > <snip> Note that we'd rather you put those outputs in your email, not in weblinks that may go away. Plus if the person who wants to help isn't always online, or they have flaky internet ... It's just accepted that emails may be large on this list ... > > Is there a way to get back my data? Is it safe to use my array while > waiting discount on HDDs and SATA cables (in one month)? > My (not very expert) money here is on a dodgy power supply! I'd be rather worried about using the computer. If the power is failing and systemd is timing out on multiple drives, it looks like the supply is no longer providing enough oomph to bring the drives up properly. That said, I don't know why it should be erasing boot blocks and the like. OMG - I was just looking for your version of mdadm, and I noticed - Seagate Baracudas! Can you let us know (a) what version of mdadm, and (b) have you fixed the timeout problem! If you haven't done (b) then that's probably your problem! Cheers, Wol
