This tutorial will show you few examples of what can happen with Areca RAID cards when the DDP is subject to odd situations or manipulation mistakes.
For various reasons you might end up with degraded, splitted and duplicated RAID sets.
In such situations, you must be very careful with the manipulations you are doing as some mistakes can result in data loss.
Basic cases
As you can see in the screenshot below, on that example one drive is missing and the state of the RAID set is not displayed. This situation can happen when the DDP crashes after a disk failure. After restarting the DDP, the RAID set state will be shown as incomplete so the corresponding logical volume will not appear in the DDP GUI.
The procedure in order to get all the volumes back is the following:
In that example the disk that is labeled as missing needs to be taken out and replaced by a new one.
Then just login to the corresponding Areca RAID card interface and do the following:
- go to the ‘RAID Set Functions’ menu
- choose the ‘Activate Incomplete RAID Sets’ function
- activate the corresponding RAID set
After that manipulation, please reboot the DDP and check that all the volumes are appearing in the DDP GUI.
In the past, a feature called “auto-activate RAID set“ was enabled by default, Areca changed that setting as automatically activating an incomplete RAID set can be dangerous. When the “auto activate RAID set” option is enabled, the RAID card automatically allocates any available disk to the incomplete RAID set and starts rebuilding the data, the problem is that under certain circumstances disks from other incomplete RAID sets might be used, resulting in data loss.
This is the reason why a manual activation of an incomplete RAID set is considered as safer. Please note that such situations should not happen if the volume check feature is used on a regular basis (at least one each 3 months).
It might happen that a RAID set gets incomplete because more than two drives failed inside a RAID 5. When this happens you have no other choice than reactivating the RAID set with the same drives in order to avoid data loss and get back the data.
In order to achieve this you will first need to login to the RAID card interface and activate the incomplete RAID set as it is explained in this tutorial, then you will have to reboot the DDP so that all the volumes get accessible again in the DDP GUI.
After you made sure that the RAID set is available, you of course need to take care of the drives that failed once by replacing them and running a volume check.
More complicated cases
The most important thing when odd cases are coming up is not to panic as inappropriate manipulations might cause data loss. We are especially advising you against using the ‘rescue RAID set’ without being explicitly told by our team to use it.
When drives are being pulled out while the DDP is powered down
The screenshot below shows an example of what can happen when one drive is pulled out while the DDP is powered off.
As you can see, one of the RAID set has been splitted into two different RAID sets with the same name.
You can easily guess that for the current situation, the only one drive contained in the duplicated RAID set should go to the one where one drive is missing.
In that case just login to the corresponding Areca RAID card interface and do the following:
- go to the ‘RAID Set Functions’ menu
- choose the ‘Activate Incomplete RAID Sets’ function
- activate the RAID set containing 7 drives out of 8
- choose the ‘Delete RAID set’ function
- delete the RAID set containing 1 drive out of 8
Then pull out the disk that is displayed as ‘free’ in the ‘system information’ page and insert it back to start rebuilding the data. After that the rebuilding completes, please reboot the DDP and check that all the volumes are appearing in the DDP GUI.
When the DDP is being rebooted while a raid set is rebuilding the data with a hot spare drive
In the screenshot below you can see a similar situation at the difference that no drive is missing in any of the RAID sets. This is why all the RAID sets states are displayed as ‘normal’.
The reason why no drive is missing is that a so-called ‘hot spare’ drive has been previously defined so that when a drive fails, this hot spare drive can be automatically used in place of the one that failed for rebuilding the data.
The problem is that the DDP has been rebooted in the middle of the operation, so when the rebuilding process completed, the raid card got confused with the failed disk and put it in a duplicated RAID set.
So in that situation as always just login to the corresponding Areca RAID card interface and do the following:
- choose the ‘Delete RAID set’ function
- delete the RAID set containing 1 drive out of 7
After you replaced the failed drive by a new one, in that specific situation you would also need to add it a hot spare as the previous one is now part of one of the RAID set.
When the DDP master head is being rebooted while a raid set is rebuilding the data with a hot spare drive
The screenshot below shows a particular case in the sense that it is the result of a sequence of wrong manipulations made on a redundant system composed of two DDP heads:
What happened originally is that one drive failed, degrading the “R1″ raid set.
A previously configured hot spare drive then automatically took the place of the failed disk and the raid card used it for rebuilding the data.
In that case just login to the corresponding Areca RAID card interface and do the following:
- Delete the TOS1 raid set that contains only 1 drive
- Delete the R6 raid set that contains only 1 drive
- Reconfigure the disks slots 10 09 and 20 17 as hot spare drives so that it can be used for rebuilding the data of an incomplete raid set
After doing that, one of the hot spare drive just took place in the incomplete raid set and the data started rebuilding.