Drive clicking, but no known errors.

  • I have a ZFS Raid Z2 array with 8 SAS drives. For over a year now, it seems like at least 1 if not 2 of the drives are bad. However ZFS reports no errors. The array is not slow, and everything works as it should. The only indication any of the drives are bad is the clicking, and the indicator on the drive caddy itself (Details below)

    Because these are SAS drives, no smart data is reported, and I am not familiar with any way to check their health status. The system is set up on a ProLiant ML350p Gen8, and each of the caddies has a central orange light that apparently blinks if the system detects issues with the drives, and 2 of the 8 drives are flashing there.

    So, should I be concerned? Is there a way to test these drives in place, or get it to give me some kind if report on them? Can I tell ZFS to do an "Extensive test" to see if it can find anything? And most intriguing, I asked 3 questions (4 if you count this one) will people actually answer all 3 of them, or just select one to answer..

  • I can't advise you on the HP system or dealing with ZFS as I don't use either, but if the system is flashing an error on 2 drives and they are clicking, I would probably first reseat the drives in the caddies and if it persists, I would replace the drives. Clicking is a mechanical noise meaning the heads are likely banging around inside the drive. It could be caused by bad connections/mating in the drive caddy, but if that doesn't fix it, you almost certainly have drives starting to fail. I've also seen tarnished connections on the under side of the drive electronics boards cause this kind of thing (particularly in humid or ocean/salt air environments), but unless you feel comfortable removing that board and trying to clean the contacts, drive replacement is the next logical step.


    RAIDZ2 will allow for 2 to fail before you loose data, and you have 2 failing drives so you are "walking a tightrope without a safety net"


    And I hope I sufficiently answered. I said what I can't answer and answered what I could.

    Asrock B450M, AMD 5600G, 64GB RAM, 6 x 4TB RAID 5 array, 2 x 10TB RAID 1 array, 100GB SSD for OS, 1TB SSD for docker and VMs, 1TB external SSD for fsarchiver OS and docker data daily backups

  • If I had even the slightest error from the array, I would without thinking twice. I was just wondering if anyone that had some experience with ZFS could speak on this.

  • As I said drive clicking is mechanical problems. It usually will not be detected by any kind of smart scan unless it is causing data errors, but it is indicative of a drive that is about to fail mechanically.


    35 years of experience dealing with drives (both singles and in large raid arrays). Clicking, unless caused by those bad connections I mentioned is always a drive with it's days numbered.

    Asrock B450M, AMD 5600G, 64GB RAM, 6 x 4TB RAID 5 array, 2 x 10TB RAID 1 array, 100GB SSD for OS, 1TB SSD for docker and VMs, 1TB external SSD for fsarchiver OS and docker data daily backups

  • If you are confidant that a clicking drive can function without causing errors I will look into replacing them (After cleaning contacts and such) I just assumed if the drive was that bad I would have had errors, or slow access, or something.
    Thank you.

  • If you don't have any data errors cleaning contacts and reseating will not hurt, but no I am not confident that a clicking drive will not cause errors, meaning that if the clicking persists after the cleaning and reseating, I would replace the drives as soon as you can. Even if the clicking does not persist, I would plan for replacement, but it may not be quite as big an emergency, so you may not have to rush to get drives ASAP, but sooner rather than later is still recommended for safety sake and piece of mind.


    However, there is a slight possibility that once the drive is powered down it may not spin up again, depending on what the mechanical issues are, so if your hardware permits hot swapping, I would try that one drive at a time. If the first one spins up again and seems to work in your array, move on to the second one. be prepared to replace the drives ASAP though, as the reseating only is on the low end of success in stopping clicking, removing and cleaning the underside board contacts will increase the chances if you are suffering from contact tarnish (putting on my electronics technicians hat, one of the few diplomas and work experience I have, the problem is not only tarnish from environmental issues, but also a film that can build up due to electro-chemical interactions as electricity flows between the contact if the contacts are not the exact same chemical composition). (Just to relate a story, Many years ago, I had an old seagate baracuda drive that was clicking. The cleanings fixed the clicking problems for a few hours, but they returned. In an attempt to recover the data, I removed the drive board and put a thin bed of solder on the contacts, so I could hopefully ensure a stable connection for data retrieval. This worked and I was able to keep the drive alive for data retrieval, but the drive continued to work without clicking for several months before the clicking returned, which is not a shocker considering the difference in the copper contacts and the lead-tin alloy solder used to make those beds without an actual solder bond. I relay this story just to hopefully let you know that this is a problem I have seen before and have dealt with).


    Reseating the drives in the caddies causes the contacts to scrape off tarnish and that electro-chemical film, but the underside of the drive boards rely on spring contacts that have no scraping action when removing and replacing the board. The best way to clean those contacts of the film is with an electronics solvent like contact cleaner or flux remover, or if you don't have those a pink pencil eraser like is on the end of a normal old wooden pencil is just as good, if not better, as it is just abrasive enough to remove the film and tarnish, where the solvents often don't get the tarnish. When doing this as a last attempt to keep a drive working, even if it's only long enough to do data recovery or cloning, I will usually gently use the eraser, and then a little contact cleaner just to ensure the contacts are as clean as possible.

    Asrock B450M, AMD 5600G, 64GB RAM, 6 x 4TB RAID 5 array, 2 x 10TB RAID 1 array, 100GB SSD for OS, 1TB SSD for docker and VMs, 1TB external SSD for fsarchiver OS and docker data daily backups

    Edited 2 times, last by BernH ().

Participate now!

Don’t have an account yet? Register yourself now and be a part of our community!