Synology crashed disk this morning



  • Always fun to walk in the office and basically hear scary beeps coming from somewhere. Is it a server? Workstation? Battery backup?

    Turns out it was the Synology DS216+. This has two 3TB drives in RAID0.
    All the lights were off but it was beeping regularly. Pressing the power button didn't do anything. I tried to access and it worked fine, must have just been asleep? All the lights came back on but disk 2 was orange.

    I was able to log in to DSM where it said disk 2 had "crashed" and I should try to backup my files.

    This was already confusing. It's a RAID0, if one drive was crashed I'd think it would be total data loss, but I could still access the folder shares just fine, hmm. SMART listed it as "normal" as well.

    Then I saw in the HDD/SSD section that disk2 had a "bad sector count" of 1. Apparently it found one lonely little bad sector and I assume marked it to avoid in the future as drives are supposed to do. But why did everything have to crash and burn over one stinking bad sector? I was hoping not to have to replace a drive for this.

    I did some reading and found out that when DSM marks something as "crashed" it may disconnect it. Since I wasn't using a redundant RAID type but could still access data, I doubt that happened, but why not try?

    I simply turned the unit off, took out both drives, sprayed some canned air while at it, and reinserted the drives.

    After turning back on, it was no longer crashed, both drives showed "normal" and the RIAD0 self-tested as healthy. It did suggest that I do an additional data validation scan which it said would be about 25 minutes per 1TB of space. I said ok, but I guess it finished in about 2 minutes while rebooting cause I didn't see this happen.

    The final message I got was that there was something wrong with a SCSI LUN something or other, I wish I took a screen shot. I'm not using the iSCSI LUN features and DSM says there is no iSCSI LUN in the system, so not sure how that message was relevant.

    In any case, I'm writting this for the future. If your Synology is beeping and talks of crashing drives and end of the world events, just stay calm, it may just be a bad sector and it's throwing a hissy fit.
    If SMART says your drive is fine, try turning it off and reseat the drives and turn back on.

    You might even remove the drive and run a test from another computer to look for further sector issues, then put it back in.

    Also I'm writing this to see if this has happened to anybody else and if I should be worried about this one bad sector issue. Drives develop bad sectors all the time, I don't know why the entire thing flips out and dies over it. I'm disappointed how the Synology handled something so simple as marking a bad sector and moving on.



  • If the drive is to the point of showing bad sectors to the Synology unit, the drive should be considered defective. It's run out of the factory assigned spare sectors, and is now having to dip into actual storage space to spare out sectors. If it's under warranty, time for a warranty exchange. If it's not under warranty, time to think about it's replacement.

    Being that it's RAID0, it shouldn't be important data anyway, right? (Something like backup staging area?)



  • @travisdh1 said in Synology crashed disk this morning:

    If the drive is to the point of showing bad sectors to the Synology unit, the drive should be considered defective. It's run out of the factory assigned spare sectors, and is now having to dip into actual storage space to spare out sectors. If it's under warranty, time for a warranty exchange. If it's not under warranty, time to think about it's replacement.

    Being that it's RAID0, it shouldn't be important data anyway, right? (Something like backup staging area?)

    The drive is only about 3 or 4 months old as far as when I actually bought it. Not sure how old it is on the store shelf!

    Is there a way to verify this somehow? Does DSM have a test feature that will prove a failure for purposes of warranty replacement? I hate that a drive MEANT for NAS boxes is showing failure in only 3 months of barely any use.
    How common is it to see a failure this soon?

    Also, you're telling me that because DSM knows there is a bad sector, the drive has completely filled up it's normal list of bad sectors? Not sure what you mean here. Some bad sectors Synology wouldn't otherwise be aware of?



  • @guyinpv said in Synology crashed disk this morning:

    @travisdh1 said in Synology crashed disk this morning:

    If the drive is to the point of showing bad sectors to the Synology unit, the drive should be considered defective. It's run out of the factory assigned spare sectors, and is now having to dip into actual storage space to spare out sectors. If it's under warranty, time for a warranty exchange. If it's not under warranty, time to think about it's replacement.

    Being that it's RAID0, it shouldn't be important data anyway, right? (Something like backup staging area?)

    The drive is only about 3 or 4 months old as far as when I actually bought it. Not sure how old it is on the store shelf!

    Is there a way to verify this somehow? Does DSM have a test feature that will prove a failure for purposes of warranty replacement? I hate that a drive MEANT for NAS boxes is showing failure in only 3 months of barely any use.
    How common is it to see a failure this soon?

    Fairly common. Which model of drive is it exactly? I ask because 'NAS' drives are really the same hardware inside, but with TLER turned on in the firmware. Which is actually kinda worthless for a RAID0. Being only 3 months, you're still in the early stages where failures are common. Almost all drives have a "bathtub curve" where failure rates are more common in the first few months, become very rare for most of the expected lifetime, and then start failing as hardware wears out. BackBlaze give you some nice graphics explaining it.

    Also, you're telling me that because DSM knows there is a bad sector, the drive has completely filled up it's normal list of bad sectors? Not sure what you mean here. Some bad sectors Synology wouldn't otherwise be aware of?

    I'd have to check the SMART stats to be certain, but yes. The drive exhausts it's reallocatable sectors and will then start reporting them to the system. I know I've seen this behavior with WD Red, Green, and Blue drives at least.



  • @travisdh1

    I guess the question is, does the sucker really need replaced? I mean really? Is this considered a "bad" drive and be covered by a warranty? I have my doubts. If I tell them "DSM says it has a bad sector, can you replace it?" They might laugh at me.



  • @guyinpv said in Synology crashed disk this morning:

    @travisdh1

    I guess the question is, does the sucker really need replaced? I mean really? Is this considered a "bad" drive and be covered by a warranty? I have my doubts. If I tell them "DSM says it has a bad sector, can you replace it?" They might laugh at me.

    Yes. Replace it. Unless your data means nothing. In which case, why have the data.



  • @guyinpv said in Synology crashed disk this morning:

    @travisdh1

    I guess the question is, does the sucker really need replaced? I mean really? Is this considered a "bad" drive and be covered by a warranty? I have my doubts. If I tell them "DSM says it has a bad sector, can you replace it?" They might laugh at me.

    Why would use use RAID0 and then question if the drive needs to be replaced?



  • @JaredBusch said in Synology crashed disk this morning:

    @guyinpv said in Synology crashed disk this morning:

    @travisdh1

    I guess the question is, does the sucker really need replaced? I mean really? Is this considered a "bad" drive and be covered by a warranty? I have my doubts. If I tell them "DSM says it has a bad sector, can you replace it?" They might laugh at me.

    Yes. Replace it. Unless your data means nothing. In which case, why have the data.

    This

    Why would you even want a somewhat questionable drive?



  • @IRJ said in Synology crashed disk this morning:

    @JaredBusch said in Synology crashed disk this morning:

    @guyinpv said in Synology crashed disk this morning:

    @travisdh1

    I guess the question is, does the sucker really need replaced? I mean really? Is this considered a "bad" drive and be covered by a warranty? I have my doubts. If I tell them "DSM says it has a bad sector, can you replace it?" They might laugh at me.

    Yes. Replace it. Unless your data means nothing. In which case, why have the data.

    This

    Why would you even want a somewhat questionable drive?

    Because it's normal for drives to have bad sectors. They get marked and life goes on. If basic and advanced SMART scans continue to say the drive is normal, and even manufacturer utility tests say it's good, how am I supposed to warranty it? Some tool has to declare the drive is bad.



  • @DustinB3403 said in Synology crashed disk this morning:

    @guyinpv said in Synology crashed disk this morning:

    @travisdh1

    I guess the question is, does the sucker really need replaced? I mean really? Is this considered a "bad" drive and be covered by a warranty? I have my doubts. If I tell them "DSM says it has a bad sector, can you replace it?" They might laugh at me.

    Why would use use RAID0 and then question if the drive needs to be replaced?

    Huh?



  • @guyinpv said in Synology crashed disk this morning:

    @IRJ said in Synology crashed disk this morning:

    @JaredBusch said in Synology crashed disk this morning:

    @guyinpv said in Synology crashed disk this morning:

    @travisdh1

    I guess the question is, does the sucker really need replaced? I mean really? Is this considered a "bad" drive and be covered by a warranty? I have my doubts. If I tell them "DSM says it has a bad sector, can you replace it?" They might laugh at me.

    Yes. Replace it. Unless your data means nothing. In which case, why have the data.

    This

    Why would you even want a somewhat questionable drive?

    Because it's normal for drives to have bad sectors. They get marked and life goes on. If basic and advanced SMART scans continue to say the drive is normal, and even manufacturer utility tests say it's good, how am I supposed to warranty it? Some tool has to declare the drive is bad.

    That is why you have high availability on anything important.



  • @IRJ said in Synology crashed disk this morning:

    @guyinpv said in Synology crashed disk this morning:

    @IRJ said in Synology crashed disk this morning:

    @JaredBusch said in Synology crashed disk this morning:

    @guyinpv said in Synology crashed disk this morning:

    @travisdh1

    I guess the question is, does the sucker really need replaced? I mean really? Is this considered a "bad" drive and be covered by a warranty? I have my doubts. If I tell them "DSM says it has a bad sector, can you replace it?" They might laugh at me.

    Yes. Replace it. Unless your data means nothing. In which case, why have the data.

    This

    Why would you even want a somewhat questionable drive?

    Because it's normal for drives to have bad sectors. They get marked and life goes on. If basic and advanced SMART scans continue to say the drive is normal, and even manufacturer utility tests say it's good, how am I supposed to warranty it? Some tool has to declare the drive is bad.

    That is why you have high availability on anything important.

    OK but that's not relevant.

    This thread isn't about merits of backup strategies or opinions on RAID0, etc.

    I'm wondering if other Synology users see DSM listing bad sectors, whether it's normal, how many are acceptable, and so on. I've even read some places that Synology can create the bad sectors via it's own reading/writing errors and they aren't "really" bad sectors at all.
    Some people take such drives and do a wipe on another computer and reinstall to Synology and it's back to normal, no bad sectors.
    Some people I've read are still using drives with over 3,000 bad sector count! Some others have hundreds or over a thousand of their own.

    Opinions range from "replace immediately" to "format it, test it and reuse". Some people say to trust the advanced SMART scan and some say SMART means nothing. Some say 1 bad sector means the internal supply of spare sectors is out and this is the beginning of the end, some say it's just fine and normal.

    What I'm looking for is actual experience or official recommendations from Synology. They don't seem to have a public post about it so I may have to do a support ticket to get answers. I'm wondering if anybody else here has drives in their Synology that show bad sectors. Plenty of people I've read continue to use drives with a count >1 and don't report issues.



  • @guyinpv Did the Synology come with the drives already installed? If so I'd ask support. If they didn't come with the Synology I'd ask the vendor support.

    What is the reallocated sector count in the SMART statistics?



  • @travisdh1 said in Synology crashed disk this morning:

    @guyinpv Did the Synology come with the drives already installed? If so I'd ask support. If they didn't come with the Synology I'd ask the vendor support.

    What is the reallocated sector count in the SMART statistics?

    It's actually performing the advanced SMART tests right now, at 40% complete. I'll know in a little bit the final details.



  • @guyinpv said in Synology crashed disk this morning:

    @DustinB3403 said in Synology crashed disk this morning:

    @guyinpv said in Synology crashed disk this morning:

    @travisdh1

    I guess the question is, does the sucker really need replaced? I mean really? Is this considered a "bad" drive and be covered by a warranty? I have my doubts. If I tell them "DSM says it has a bad sector, can you replace it?" They might laugh at me.

    Why would use use RAID0 and then question if the drive needs to be replaced?

    Huh?

    It was a bit tongue in cheek, RAID0 provides no protection against a drive failure, one drive has partially failed. So you've likely lost the data that was in that section of the disk.

    And you've asked if you should retry the drive, or replace the drive.



  • @DustinB3403 said in Synology crashed disk this morning:

    @guyinpv said in Synology crashed disk this morning:

    @DustinB3403 said in Synology crashed disk this morning:

    @guyinpv said in Synology crashed disk this morning:

    @travisdh1

    I guess the question is, does the sucker really need replaced? I mean really? Is this considered a "bad" drive and be covered by a warranty? I have my doubts. If I tell them "DSM says it has a bad sector, can you replace it?" They might laugh at me.

    Why would use use RAID0 and then question if the drive needs to be replaced?

    Huh?

    It was a bit tongue in cheek, RAID0 provides no protection against a drive failure, one drive has partially failed. So you've likely lost the data that was in that section of the disk.

    And you've asked if you should retry the drive, or replace the drive.

    Well that's what threw me off. Even with the error and the message that it was "crashed", I could access all the files just fine. The shared folders were there, etc. It was even telling me to copy the files off just in case. So on one hand it said there was a crash, on the other hand, nothing was broken.

    Many people said they just took out the drive and reinserted it and all was well. This is what I did and it came back to life, verified the RAID and so the only issue is that it now lists one bad sector on drive 2.



  • @guyinpv said in Synology crashed disk this morning:

    @DustinB3403 said in Synology crashed disk this morning:

    @guyinpv said in Synology crashed disk this morning:

    @DustinB3403 said in Synology crashed disk this morning:

    @guyinpv said in Synology crashed disk this morning:

    @travisdh1

    I guess the question is, does the sucker really need replaced? I mean really? Is this considered a "bad" drive and be covered by a warranty? I have my doubts. If I tell them "DSM says it has a bad sector, can you replace it?" They might laugh at me.

    Why would use use RAID0 and then question if the drive needs to be replaced?

    Huh?

    It was a bit tongue in cheek, RAID0 provides no protection against a drive failure, one drive has partially failed. So you've likely lost the data that was in that section of the disk.

    And you've asked if you should retry the drive, or replace the drive.

    Well that's what threw me off. Even with the error and the message that it was "crashed", I could access all the files just fine. The shared folders were there, etc. It was even telling me to copy the files off just in case. So on one hand it said there was a crash, on the other hand, nothing was broken.

    Many people said they just took out the drive and reinserted it and all was well. This is what I did and it came back to life, verified the RAID and so the only issue is that it now lists one bad sector on drive 2.

    My point is, if you care about the data, don't use RAID0.



  • A single drive would be safer compared to RAID0, because with RAID0 if a drive does go, you lose all of the data stored on that drive.

    Guessing that this is a 2-drive unit you have what 8TB max in this unit (2x4TB raid0)?

    Just purchase a single 8TB drive.



  • @DustinB3403 said in Synology crashed disk this morning:

    A single drive would be safer compared to RAID0, because with RAID0 if a drive does go, you lose all of the data stored on that drive.

    Guessing that this is a 2-drive unit you have what 8TB max in this unit (2x4TB raid0)?

    Just purchase a single 8TB drive.

    A single drive is not safer because if the drive goes you still lose all of the data on that drive.



  • @stacksofplates said in Synology crashed disk this morning:

    @DustinB3403 said in Synology crashed disk this morning:

    A single drive would be safer compared to RAID0, because with RAID0 if a drive does go, you lose all of the data stored on that drive.

    Guessing that this is a 2-drive unit you have what 8TB max in this unit (2x4TB raid0)?

    Just purchase a single 8TB drive.

    A single drive is not safer because if the drive goes you still lose all of the data on that drive.

    It is because in raid 0 any one of the drives die you lose all the data... 2 drive raid, twice as likely for a failure.



  • @stacksofplates said in Synology crashed disk this morning:

    @DustinB3403 said in Synology crashed disk this morning:

    A single drive would be safer compared to RAID0, because with RAID0 if a drive does go, you lose all of the data stored on that drive.

    Guessing that this is a 2-drive unit you have what 8TB max in this unit (2x4TB raid0)?

    Just purchase a single 8TB drive.

    A single drive is not safer because if the drive goes you still lose all of the data on that drive.

    With a single drive, you have a lower chance of failure (raid controllers, drive issues whatever) And a single drive is easily copied from for backup purposes.

    You get an equal size drive, and robocopy or DD all of the data.

    Backup done.



  • @brianlittlejohn said in Synology crashed disk this morning:

    @stacksofplates said in Synology crashed disk this morning:

    @DustinB3403 said in Synology crashed disk this morning:

    A single drive would be safer compared to RAID0, because with RAID0 if a drive does go, you lose all of the data stored on that drive.

    Guessing that this is a 2-drive unit you have what 8TB max in this unit (2x4TB raid0)?

    Just purchase a single 8TB drive.

    A single drive is not safer because if the drive goes you still lose all of the data on that drive.

    It is because in raid 0 any one of the drives die you lose all the data... 2 drive raid, twice as likely for a failure.

    Ah ya fair enough. I was just thinking you still only need to lose one either way. It's been a long day.



  • Turns out neither of the two drives have a reported bad sector in SMART. The only thing I can see is that disk 2 has SMART value "current_pending_sector" of 1. Apparently this means there was a write or read issue and it will be marked the next time it attempts to be accessed.

    As far as I've researched, having 1 pending sector issue is nothing to worry about.

    I just wonder why it didn't handle it automatically and further, wonder if it's striping that somehow prevented it from auto-fixed the sector. I may just have to use these drives separately and skip the striping.



  • @guyinpv i would agree. Nothing to worry about.


Log in to reply