Home
Join
check
  • spicehead-eacgo wrote:

    So I desactivated the disk that has a problem, turned off the NAS, replaced the disk and after chose the option REPAIR for the disk in conserne. Now it is in the state of reparing ( this tuto I watched on internet).

    I don't know if I did it correctly, but that's what I can do for the moment.

    As little as I understood, you did not make a full backup before replacing the disk. So you're almost right with your conclusion do be unable to do more for the moment. You may pray that no other disk of your array fails before the rebuild resp. repair finishes. If all disks are of same age, there is high risk that another one fails. But if these are of different age, you may have luck and succeed repairing.

    And as you'll see the repair time, that's why you may follow the recommendation to switch to two disk failure tolerance AFTER the repair finished. Such a change of hybrid RAID type is builtin into your Synology, as reported in the previously reported Synology article.

    I don't know how your backup concept is currently as I did not see it mentioned. As you're new, you may not know yet neither and have to ask your colleagues. If you're in the better position, your predecessor has established a sound backup concept and your company is doing sound and routine automated backup. You may check if this is indeed the case and if it meets the needs of your company. But Anhalt fears that the lack of mentioning backup constraints might be an indication that backup either does not exist or does not meet business needs, although expressing it differently. I share such fears. And his proposal to use an external hard disk is not make it a sound solution although an improvement. And in that case, you would want to establish a sound backup and restore concept, aligned with business needs. There already exist good tutorials if you're not familiar with current good options. You also may find some on Synology knowledge base as well as in some how-to or article on our forum (I didn't check but know that at least solved topics exist here on that backup subject).

    spicehead-eacgo wrote:

    Here they told me that I should not do any changes about type of RAID, cause of my experience, but that I need to replace the disk.

    The argument about RAID type change is not correct, especially as you're willing to learn. There argument about order is correct. You don't want to change RAID type while the RAID is degraded. And after you got a healthy RAID, you'll want to change disk failure tolerance to two disks when your disks are spinning hard disks. This is possible also with your current hybrid RAID as reported in the previously referenced Synology article. And as I don't know the current level of usage of your current RAID array, I don't know when other changes become appropriate, like increase of net storage capacity or increase of volume for snapshot capacity.

    ranhalt wrote:

    You're worried that replacing the bad drive will cause data loss

    That's my worry now, not the OPs worry. That's why I recommended a full backup before starting disk replacement. He may start to be concerned when reading of next disk failure DURING repair.

    Was this post helpful? thumb_up thumb_down
  • View Best Answer in replies below

    16 Replies

    • I have never done this specific thing on a Synology NAS but have had a couple of hiccups (ranging from what are the implications of skipping firmware versions to what does this purple light mean) and have always had awesome success reaching out directly to Synology support.  

      Pepper graySpice (1) flagReport
      Was this post helpful? thumb_up thumb_down
    • DS418 drives are hotswappable.

      I replaced a bad drive in a 418 recently.  After replacing the drive, I had to go into the management interface and tell it to resync the array - I forget the exact terminology used.

      Pepper graySpice (4) flagReport
      Was this post helpful? thumb_up thumb_down
    • I'm not sure why you think that device isn't actively supporting hot swap capabilities, it is absolutely hot swappable.

      You shouldn't even need to initiate the rebuild, you can just check the status after a few minutes of swapping the drive. If it doesn't start rebuilding, you'd just go into the array actions and initiate the rebuild.

      On the inverse, shutting it down to swap the drive cold would be a greater risk.

      Pepper graySpice (3) flagReport
      Was this post helpful? thumb_up thumb_down
    • Sorry, I didn't get if you use RAID of which type. I further didn't get the kind of disks (SSD or spinning disks) and their capacity.

      If you created no RAID, then you've to restore data from backup as only option. If you created a RAID 5 with large spinning disks, you'll likely loose all data during rebuild, that's why you want to do a full backup before replacement.

      As others have noted, I did not get what special disks you have so that they would not support hot swapping although your NAS supports hot swapping.

      Pepper graySpice (1) flagReport
      Was this post helpful? thumb_up thumb_down
    • Thank you everyone for quick answers :)

      Actually, it uses Sunology Hybrid RAID ( that supports lose of 1 disk only). I saw online that supports hot swappable, but down someone that worked before me wrote with big red letters that is NOT hot swappable.

      I don't find any video that shows how to shange disk on not hot swappable RAID.

      I am sorry for bothering you like this, but I am left all alone in this company and not so much experience so I am in complete mess.

      Was this post helpful? thumb_up thumb_down
    • spicehead-eacgo wrote:

      Actually, it uses Sunology Hybrid RAID ( that supports lose of 1 disk only).

      Thanks for partial clarification. But I still could not find the disclosure of disk types and their capacities.

      With todays common disk capacities, you want RAID 1, 10 or 6 for spinning disks of current capacities. If all your disks are SSDs, then your choice of RAID type might be appropriate although a bit risky as proprietary.

      spicehead-eacgo wrote:

      I saw online that supports hot swappable, but down someone that worked before me wrote with big red letters that is NOT hot swappable.

      I don't know what motivated your predecessor to write so. As little as I remember, all hard disks within more than the last 15 years should support hot swapping. But for hot swapping, there is some support needed on the disks themselves as well as on the side of the controller on which they are attached. NAS are typically built with controller setups with hot swap support. So if your setup is younger, then there are other reasons for such a note. This might be selection of hard disk type not recommended for NAS usage or inappropriate RAID type.

      I don't know the capacity of your RAID nor the volume of data in your RAID. I don't know how long it takes to do a full backup and a restore from backup. You might consider to do a full backup, to set up your RAID fresh without degraded disk with more appropriate RAID type and restore your backup on new RAID. There are undisclosed reasons for the current RAID type. In your preparations, you may want to remove such reasons. This may require to purchase other hard disks too before such an action. I guess that Synology also has white papers and guides for better setup and selection. (I'm a QNAP user with similar NAS. And QNAP does not offer such a hybrid RAID type. QNAP clearly recommends to not mix storage device types in same RAID. And a mix of capacities does not make much sense, except perhaps for some temporary live migration. And QNAP NAS are not as easy to manage as Synology NAS.)

      Pepper graySpice (1) flagReport
      1 found this helpful thumb_up thumb_down
    • Thank you, these are very usefull information you gave me.

      There are 4 HDD of 4TB.

      I don't know much about RAID, I was just thrown here without much information, so I need to explore. Thank you very much

      Pepper graySpice (1) flagReport
      Was this post helpful? thumb_up thumb_down
    • spicehead-eacgo wrote:

      I don't know much about RAID, I was just thrown here without much information, so I need to explore.

      So that's something you want to learn anyway as sole IT guy in your company (unless managed by 3rd party). Synology and QNAP have tutorials, white papers and guides on RAID. Other vendors have such training material too. There exist also training videos, some vendor specific and some vendor agnostic. Standard RAID types should work across vendors. I've never tried and heard that this might be more limited than expected. Definitely the different vendors support different RAID types and RAID options. Synology hybrid RAID is such a vendor specific example. For the case of QNAP, ZFS resp. RAID Z support has been extended to another of their platforms. While this RAID type exists already for longer, I don't know if and when it may become a standard RAID type. And if I remember it right, QNAP implementation on their main operating system has some limitations on supported options for RAID Z.

      And don't forget to provide feedback how you addressed your situation when you finish to implement a solution. While working with degraded RAID, backup becomes even more important. You may not know if your learning advances quick enough before another RAID failure occurs. But RAID basics should be quick enough to learn. You may postpone more advanced options (e.g. snapshots) until later.

      I saw your update after I replied. So I'm updating too. As you have 4 disks of 4 TB, seemingly spinning disks, I don't see why your predecessor selected hybrid RAID, unless some are using SMR encoding or some consumer model (e.g. for gamers or desktops) and not for 24/7 operation in a NAS. With 4 disks, RAID 10 and 6 are available options. RAID 6 supports two disk failures. RAID 10 supports 1-2 disk failures, depending where they occur. Both RAID types require a minimum of 4 storage media. RAID 10 is more performant and RAID 6 provides a higher net storage capacity. So I guess that you may prefer RAID 6. With RAID 6, you may get lower net storage capacity compared to your current hybrid RAID. That's why it is relevant to which degree your net storage capacity is already in use. With RAID 6, you'll get a net storage capacity with your current hard disks of 8 TB.

      Here is a Synology articleOpens a new window, reporting that your hybrid RAID may be converted to two-disk failure tolerance if all your disks would be healthy, see its last paragraph.

      Pepper graySpice (1) flagReport
      Was this post helpful? thumb_up thumb_down
    • Hello to everyone - again. I read your advices carefully and I appriciate you all for your time. Here they told me that I should not do any changes about type of RAID, cause of my experience, but that I need to replace the disk.

      So I desactivated the disk that has a problem, turned off the NAS, replaced the disk and after chose the option REPAIR for the disk in conserne. Now it is in the state of reparing ( this tuto I watched on internet).

      I don't know if I did it correctly, but that's what I can do for the moment.

      Thank you again

      0 of 1 found this helpful thumb_up thumb_down
    • spicehead-eacgo wrote:

      Thank you, these are very usefull information you gave me.

      There are 4 HDD of 4TB.

      I don't know much about RAID, I was just thrown here without much information, so I need to explore. Thank you very much

      https://en.wikipedia.org/wiki/Standard_RAID_levels

      I get that you're worried and you're not wrong to be for a different reason: RAID isn't a backup, just a redundancy in case a drive fails like this. You're worried that replacing the bad drive will cause data loss, but you're dealing with 4x4TB in a RAID5, which means 12TB of capacity, maybe less data than that. Do you have it in your budget to buy a 10TB external hard drive for $200 just to take a backup of the data? That might alleviate your concern and give you an emergency backup drive for cold storage. But the NAS supports hot swapping. The reason why your predecessor doesn't work there anymore is because they didn't know what they were talking about.

      Pepper graySpice (1) flagReport
      Was this post helpful? thumb_up thumb_down
    • Thank you ranhalt​ for your reply. Actually I was worried more about replacing disk on Hot swappable NAS, as I just watch video for the other way. So I read a lot of comments about losing data. But you are right that one drive of 10TB would be a good solution. Thanks

      Was this post helpful? thumb_up thumb_down
    • scheff1 wrote:

      spicehead-eacgo wrote:

      I don't know much about RAID, I was just thrown here without much information, so I need to explore.

      So that's something you want to learn anyway as sole IT guy in your company (unless managed by 3rd party). Synology and QNAP have tutorials, white papers and guides on RAID. Other vendors have such training material too. There exist also training videos, some vendor specific and some vendor agnostic. Standard RAID types should work across vendors. I've never tried and heard that this might be more limited than expected. Definitely the different vendors support different RAID types and RAID options. Synology hybrid RAID is such a vendor specific example. For the case of QNAP, ZFS resp. RAID Z support has been extended to another of their platforms. While this RAID type exists already for longer, I don't know if and when it may become a standard RAID type. And if I remember it right, QNAP implementation on their main operating system has some limitations on supported options for RAID Z.

      And don't forget to provide feedback how you addressed your situation when you finish to implement a solution. While working with degraded RAID, backup becomes even more important. You may not know if your learning advances quick enough before another RAID failure occurs. But RAID basics should be quick enough to learn. You may postpone more advanced options (e.g. snapshots) until later.

      I saw your update after I replied. So I'm updating too. As you have 4 disks of 4 TB, seemingly spinning disks, I don't see why your predecessor selected hybrid RAID, unless some are using SMR encoding or some consumer model (e.g. for gamers or desktops) and not for 24/7 operation in a NAS. With 4 disks, RAID 10 and 6 are available options. RAID 6 supports two disk failures. RAID 10 supports 1-2 disk failures, depending where they occur. Both RAID types require a minimum of 4 storage media. RAID 10 is more performant and RAID 6 provides a higher net storage capacity. So I guess that you may prefer RAID 6. With RAID 6, you may get lower net storage capacity compared to your current hybrid RAID. That's why it is relevant to which degree your net storage capacity is already in use. With RAID 6, you'll get a net storage capacity with your current hard disks of 8 TB.

      Here is a Synology articleOpens a new window, reporting that your hybrid RAID may be converted to two-disk failure tolerance if all your disks would be healthy, see its last paragraph.

      They probably selected SHR because depending on the DSM image, it's the default. I'm more perplexed why OP has completely ignored advice on here with the disk replacement, YouTube was always an option over a thread of pros.

      Pepper graySpice (1) flagReport
      1 found this helpful thumb_up thumb_down
    • spicehead-eacgo wrote:

      Hello to everyone - again. I read your advices carefully and I appriciate you all for your time. Here they told me that I should not do any changes about type of RAID, cause of my experience, but that I need to replace the disk.

      So I desactivated the disk that has a problem, turned off the NAS, replaced the disk and after chose the option REPAIR for the disk in conserne. Now it is in the state of reparing ( this tuto I watched on internet).

      I don't know if I did it correctly, but that's what I can do for the moment.

      Thank you again

      Doing what you did is a horrible idea and is the riskiest method you could have used. Nobody mentioned to do that from what I can see, everyone said do the opposite.

      Good luck, I hope your data is OK and that you have good backups.

      Pepper graySpice (2) flagReport
      Was this post helpful? thumb_up thumb_down
    • spicehead-eacgo wrote:

      So I desactivated the disk that has a problem, turned off the NAS, replaced the disk and after chose the option REPAIR for the disk in conserne. Now it is in the state of reparing ( this tuto I watched on internet).

      I don't know if I did it correctly, but that's what I can do for the moment.

      As little as I understood, you did not make a full backup before replacing the disk. So you're almost right with your conclusion do be unable to do more for the moment. You may pray that no other disk of your array fails before the rebuild resp. repair finishes. If all disks are of same age, there is high risk that another one fails. But if these are of different age, you may have luck and succeed repairing.

      And as you'll see the repair time, that's why you may follow the recommendation to switch to two disk failure tolerance AFTER the repair finished. Such a change of hybrid RAID type is builtin into your Synology, as reported in the previously reported Synology article.

      I don't know how your backup concept is currently as I did not see it mentioned. As you're new, you may not know yet neither and have to ask your colleagues. If you're in the better position, your predecessor has established a sound backup concept and your company is doing sound and routine automated backup. You may check if this is indeed the case and if it meets the needs of your company. But Anhalt fears that the lack of mentioning backup constraints might be an indication that backup either does not exist or does not meet business needs, although expressing it differently. I share such fears. And his proposal to use an external hard disk is not make it a sound solution although an improvement. And in that case, you would want to establish a sound backup and restore concept, aligned with business needs. There already exist good tutorials if you're not familiar with current good options. You also may find some on Synology knowledge base as well as in some how-to or article on our forum (I didn't check but know that at least solved topics exist here on that backup subject).

      spicehead-eacgo wrote:

      Here they told me that I should not do any changes about type of RAID, cause of my experience, but that I need to replace the disk.

      The argument about RAID type change is not correct, especially as you're willing to learn. There argument about order is correct. You don't want to change RAID type while the RAID is degraded. And after you got a healthy RAID, you'll want to change disk failure tolerance to two disks when your disks are spinning hard disks. This is possible also with your current hybrid RAID as reported in the previously referenced Synology article. And as I don't know the current level of usage of your current RAID array, I don't know when other changes become appropriate, like increase of net storage capacity or increase of volume for snapshot capacity.

      ranhalt wrote:

      You're worried that replacing the bad drive will cause data loss

      That's my worry now, not the OPs worry. That's why I recommended a full backup before starting disk replacement. He may start to be concerned when reading of next disk failure DURING repair.

      Was this post helpful? thumb_up thumb_down
    • Now you really scared me!

      Was this post helpful? thumb_up thumb_down
    • spicehead-eacgo wrote:

      Now you really scared me!

      It is not too late as long as you did not get next disk failure. You should not abort the current repair action.
      If you forgot to initiate a full backup before you replaced the failed disk, you may still initiate a backup ASAP. And if you have a valid backup of not long ago, a differential backup might take short enough time before another disk failure happens. Being prepared and having such a safety net is what we prefer. And if you don't have any backups, a full backup (without verification) may take shorter than the RAID repair time.
      Pepper graySpice (2) flagReport
      1 found this helpful thumb_up thumb_down

    Read these next...