RAID arrays are usually intended to keep important information, the loss of which is undesirable. Therefore, all who utilize RAID arrays might need to know a list of precautions about array use.
1. Are there enough coolers to cool the storage down?
Any hard drive gets hot during its operation, and if there are several of them, then negligence of cooling regulations and requirements may cause breakdowns. If the array overheats, it can fail, so you need to ensure an adequate cooling. Before configuring the storage, you need to make sure if a working cooler is in, if there is enough space around the array and to prevent its overheating.
2. Use UPS
If a power outage occurs during data writing, then the data will not be written. Moreover, there is an increased danger of logical and mechanical disk failures. In order to eliminate this, use an uninterruptable power supply (UPS).
First, it will save the equipment from the results of voltage flicker. Second, in case of power outages, the user will have time to save files and to switch off the equipment safely.
3. Take disks of different production groups
Most RAID arrays, except RAID0, are optimized for work in a mode of limited functionality. If one disk (up to two in RAID6) breaks down, the system will continue working with a reduced read/write speed.
If more than one disk (two disks) break down while working in a degraded mode, data will be lost. For this reason, it’s recommended to arrange the array from disks of different production groups. In this way, you can reduce the likelihood of concurrent disk failure.
1. Disk order may be important
Each disk in a RAID array has its sequence number and data is written and read in a certain order. Some controllers define the order of disks by the order they are connected. If the disk order is changed, depending on the configuration, there might occur some typical problems:
- RAID0: the system is not able to initialize the disk data correctly because the block sequence is changed;
- any RAID with a parity code: the system may not be able to understand the data. There is a high possibility that any further manipulation (e.g. data writing) will lead to loss of some blocks on the messed-up disks. Recovery of this data requires an additional software application and specialized knowledge.
The system of most advanced RAID arrays may define by itself, if the disk is connected in the right order. This can be helpful during diagnostics.
2. To format or not to format?
Let’s assume that the user had to eject a drive from the RAID array and to connect it to a PC for some reason (e.g. diagnostics). In such a case, it’s necessary to read operating system messages carefully.
The system may offer formatting the device. It’s not worth accepting this offer. The thing is that the formatting may lead to loss of valuable information located on the disk followed by a long and time-consuming recovery.
3. Connecting, disconnecting disks
If the data storing system does not show the possibility of a spare disk, it’s required to disconnect it from power supply before any physical operations with the RAID array. If you don’t do that, incautious actions may lead to mechanical disk damages or logical data damages.
If you use the hot spare, please, be aware of the following precautions:
- You cannot use this function for disks that are part of functioning RAID at that moment in a degraded mode;
- Do not disconnect more than one disk per one time. After RAID disk disconnection and its replacement make sure that RAID rebuild is over. Disconnect any further disk only after that.
The device transportation requires enhanced caution to protect the disks from mechanical damages. Do not throw and do not shake the device. If a long trip is waiting for the array, you should pack it in antistatic plastic and aluminum foil and protect it against vibrations using a special package.
What is not allowed to do with the RAID array: checklist
Before you start any operations with the RAID array, please, get familiar with the following checklist:
Do not connect the RAID array to the device that does not have sufficient number of coolers.
Do not change the disk order.
During the disk diagnostics, read all the operating system messages.
Before any physical operations switch the disks off from the net.
Eliminate mechanical disk damages.
Do not expose the array to vibrations and other loads during the transportation.