OK guys this is whats going on.
I have a dell poweredge 2600 with 6 HD's installed on it. Just recently we had a server crash witch lead to a reinstall of the OS. Before, i had the computer seeing all the drives separately, OS on one drive, website data on another drive, and the other 4 where blank drives.
After reinstalling the computer with the "Dell open manage server assistant" i found out that all my hard drives are being seen as one, with one partition being the OS, another being the dell utility partition, and the rest being open space.
This is my goal.
I would like the server to see all HD's separately. Have the first drive partition into 2, one partition for the "dell utility partition" and the other being the OS.
I would like to make the 2nd HD the storage for my website.
Have the 3rd HD make an exact copy of the first drive(For OS Backup)
Have the 4th HD and exact copy of the 2nd HD (For website Backup)
HD 1 (Dell Utility Partition/OS)
HD3 (Backup of HD1)
HD4 (Backup of HD2)
Basically all i need to to know how to make all the HD's separate and not as one, and how to make a mirror image of the specified HD's
I have the following software:
Dell Openmanage server assistant boot cd
Dell Openmanage system management
Windows 2000 server
If i need to to another install of the OS, that is fine. Just let me know what i need to read or do to make this happen..
Thanks for all your help 🙂
When you use the Dell installation utility, it will configure the most common RAID configuration, based on the disks in the system. In this case, a RAID 5, using all disks would likely have been the default. This will appear to Windows as a single drive. You can configure RAID first, then skip that step during the setup. Using the installation utility is the only way to get the Dell Utility Partition. Is there a specific reason you have to have the Utility Partition? You can always run diagnostics from a CD, USB, and/or network. Anyway ...
I will first tell you how to make each drive appear individually in Windows, but then I will propose another idea that will better suit your goal.
In order for a drive that is connected to a RAID controller to be seen in Windows, it has to be configured by the controller first. The only RAID level that will take a single drive is a RAID 0. So, if you set each drive up as a single-disk RAID 0, then you will see all the drives in Windows Disk Management. You cannot leave them unconfigured and have them available in Windows.
Alternatively, you could turn the RAID controller off - put it in SCSI mode, so that each hard drive is seen individually, however, as you will see in the next paragraph, this is not the best way to do what you wish to do.
Here is what I propose you do, based on your wish to have a backup ("exact copy") of the OS and Website:
Configure drives 1 and 2 as a RAID 1, and drives 3 and 4 as RAID 1, then, if you need individual drives then you can configure them as single-disk RAID 0 so they show in Windows as individual disks. However, you have in your configuration as having these drives "blank". Do you have a plan for your "blank" drives? There are many other things you could do with these drives if you don't have a plan for them.
By configuring your disks in RAID 1, an "exact copy" is made by the RAID hardware, leaving the system resources to run the OS. RAID 1 is also called "mirroring". Having them in a mirror, if one disk in the RAID 1 fails, then the array will stay running on the remaining drive as if nothing had happened (although it will of course tell you that a drive failed, but there will be no disruption in it).
Do what you like, but I would recommend configuring a RAID 1 for your OS, a RAID 1 for your Website, then configuring another RAID 1 with the remaining two drives. If you don't need any extra storage space, you can configure the other two as Hot Spares, to automatically take the place of any drive that might fail.
Thank you for your help.
I am always open to new suggestions and I think I like yours better than what I was trying to do.
I like the idea of the hot spares and will defiantly look into this.
After having this crash I’m trying to find new ways to not only prevent this from happening again, but if it does happen again to reduce the amount of down time.
Once again, I am not an expert in this field of RAID, so any ideas and suggestions would be greatly appreciated.
I am trying to read as much documentation as I can about RAID, so if you have any you would like to recommend, please do.
That sounds logical and is exactly why RAID was invented and why it is used. When you lose a drive, there IS no downtime. (EDIT: In most cases there is no downtime. In a situation where the drives are cabled instead of attached to a hot-swap backplane, then the system would have to be powered down to replace the failed drive, but at least will stay running and it can be taken down on your own terms.)
There are lots of places that will explain the basics of what RAID is, why it is used, different RAID levels, but here is a very basic run-down of the main RAID levels you will probably ever need:
RAID 0 - Performance configuration, stripes data across all disks in the array, one disk minimum is required, fast disk access because controller can read data from multiple disks at once, NO redundancy so if one drive fails then the array is gone, storage space is equal to the number of disks times the size of the smallest disk (4x500GB=2000GB).
RAID 1 - Disk mirroring, each disk holds identical data so that if one drive fails, the other will automatically - and seamlessly - take over until the other drive is rebuilt, very common RAID level for an OS volume, requires two disks, storage space is equal to the size of the smallest disk (2x500GB=500GB).
RAID 5 - Very common for data volumes when more space is required, data and parity (used to calculate missing data when a drive fails) are striped across all disks, minimum of 3 disks is required, because of parity, storage space is equal to the number of disks times the smallest disk MINUS one (n-1), one disk can fail and the array will still be operational.
RAID 10 - Very common for databases as there is good read/write performance, this is also called RAID 1+0 as it is a RAID 0 with a RAID 1 for each leg of the RAID 0, minimum of 4 disks is required and must contain even numbers of disks, storage space is equal to the number of drives times the size of the smallest drive DIVIDED by 2 (an expensive configuration), the array will stay operational with up to two drives failed (in a 4-disk config) - as long as you don't have both members of a RAID 1 fail.
Dell PERC 6 and newer/higher can also do RAID 6 and 60, but I don't know why anyone would use them (RAID 6 is RAID 5 with the ability to lose two disks, storage space is equal to n-2).
You can read about these and more at the link below, or go to wikipedia or webopedia for even more information.
There are a lot of real-world information about RAID that is hard to come by without experience (rebuilds, stripe size, reconfigures, errors). If you have questions beyond the basics of most reference material, or if you encounter problems or questions while working with a RAID array, the best thing to do is ask an expert. Here is a good place to find answers to those types of questions/problems.
To set things up like this, you will need to reinstall, go to CTRL-M and configure drives 1 and 2 (they are actually labeled drives 0 and 1) as a RAID 1, Initialize the array, then install the OS. Once installed, you can use Dell's management utility to set up all the remaining drives.
I fully understand what you are saying, and the logic of setting it up. But what i lack is the knolege of how to use this program once you press Crtl+M. could you give me some detaild information on how to use this program or tutorials.
This would be usefull information.
If you can just get the to the starting line i can figure it out from their.
I know that you have given me more information than i have asked for and i am truly happy for this, in the last few minutes i have learned a ton on information about RAID. I just dont know how to use this program.
PERC/CERC BIOS Configuration Utility U827
Thank you so much for everything..
(EDIT) I Figured out how to make the HotSpare. I just have not figured out how to move the drives 0 an 1 from raid ch0 to ch1. and so on.
it seems that drives 0,1,2 are on raid ch0 and drives 3,4,5 are on ch1.
How do i move the drives around?
I believe you have the daughterboard installed that "splits" the backplane into two channels. You can take it out, but it really doesn't make any difference ... the controller can manage drives on different channels in the same array. If you didn't want to split array drives across channels, you could do a RAID 1 with 0,1 and set 2 as its HS, then 3,4 as RAID 1 with 5 as its hotspare.
As far as using CTRL-M ... you will go to Configure, View/Add Configuration to create an array (Configure, Clear Configuration to clear - before you create). Once it shows you the drives, hit Space Bar to select (it will start to blink) all the drives that will be in the array, then hit Enter (they will stop blinking), then hit F10, then Space Bar to select the Logical Disk to be created and F10 again. Select your options, like RAID 1 (don't change Stripe Size or other default settings unless you know what you're doing). Then Accept. After the message about initializing newly created Logical Drives, then go to Initialize on the main menu, select the Logical Drive by hitting Space Bar, then F10 to Initialize. Here is a guide on setting up an array:
Let us know if you have specific questions along the way.
"Select your options, like RAID 1 (don't change Stripe Size or other default settings unless you know what you're doing). "
This is where i got stuck, My server is only showing RAID level 0 and 5. no raid level 1
Question, when installing the OS, Should i install from the dell server assistant cd or from the Original Microsoft Server 200 cd. If i do need to install from the original cd, could you direct me to the raid drivers.
EDIT, i also have a blinking orange light above my drives, is this bad?.