70-210 - basic or dynamic disk

Author

basic or dynamic disk

ne0-reloaded

2004-01-20, 12:06 am

im working on my 70-210 and gotta question about basic and dynamic disk. is there an advantage to using one over the other? also in respect to dynamic disk, if i have two hdd of 2 diff sizes and speeds, can i use spanned or striped volumes to make them into one logical drive? im confused about these topics so add input if u can.

thanks

aznluvsmc

2004-01-20, 10:57 am

The advantage of using dynamic disks are:

Ability to dynamically expand volumes without rebooting

Support for software RAID implementations

I'm not sure about using drives of different speed and size to implement disk spanning because the best practice is to use disks of the same type.

In all honesty, I don't think many servers implement dynamic disks as hardware RAID is used more often.

pcthug

2004-01-21, 5:39 pm

If a network administrator uses dynamic disks in his enterprise or mid-size network production servers, he must be smoking crack cocaine. Dynamic disk is best for workstations. Hardware RAID is the way to go, if you can afford it. It gives you the "real" fault tolerance as opposed to the virtual RAID based on dynamic disk.

To strip a dynamic disk, other disks must have the same size of unallocated spaces. To span, size doesn't matter as long as you have any size of unallocated free space on other disks.

Remember:

1) All the basic primary and extended partitions become simple volumes in dynamic disks.

2) You can't expand the boot or system volumes in dynamic disks.

3) In order to expand any simple volume, it must've been created after you converted the disks to dynamic disks.

4) You can only have simple, expanded, stripped and mirrored volumes in windows 2000 pro.

5) You can't install windows 2000 on simple volumes that are created after the dynamic disk conversion. The simple volume must contain FAT or MFT, which are only available in basic disk configuration. So, any basic partitions that became dynamic simple volume can be used for win2k installation.

5) If any one of the disks in the array fails, your data are hosed except with the mirrored volumes, which let you recover the data from one of its mirrors. Mirror won't help, if the IDE or SCSI controllers fail.

6) You can span or strip up to 32 disks.

7) You can covert dynamic disks arrays back to basic, as long as you remove the spanned, stripped and mirrored volumes apart.

8) Only the server supports RAID 5 and disk duplexing. You need at least three disks for RAID 5 and the size of all the disks must be the same. The disk speed doesn’t have to be the same, although not recommended. The final size of the Raid 5 array is equal to the entire size of the array minus one of the disk size. For example, if you have three 30GB disks and want to implement Raid 5, the final size of the array will be 90GB-30GB=60GB. Sometimes, this rule is called 1/3 rule. The reason for this is that one disk will be used for parity information. If one of the disk crashes, a server will continue to work without any problem, but you must replace the crashed disk immediately. RAID 5 will automatically recreate data on the new disk. You can go further by mirroring the Raid 5 array. This is called RAID 10, which you don’t have to worry about for 210 or 215.

Disk duplexing is similiar to mirroring except disks must be on different controllers. For example, if you have a single IDE channel and two disks on it as master and slave, you can only mirror them. However, if you have one disk on the IDE channel 0 and an another on the IDE channel 1, you can have the disk duplex.

Sponsored Links