System and method of volume group creation based on an automatic drive selection scheme
View Patent ↗A method and a system for volume group creation based on an automatic drive selection scheme are disclosed. In one embodiment, a method of a redundant array of independent drives (RAID) includes processing at least one attribute data describing a volume group of the RAID. The method also includes automatically selecting a plurality of drives of the volume group based on at least one of an optimum tray distribution of the volume group, a spindle speed of each drive in the volume group, and an alignment of the plurality of drives in the volume group. In another embodiment, a system of a RAID includes a user device to communicate an attribute data describing a volume group of the RAID, and a drive selection module to automatically select a plurality of drives of the volume group based on the attribute data using a baseline selection algorithm.
1. A method of a redundant array of independent drives (RAID), comprising:
processing at least one attribute data describing a volume group of the RAID, wherein the at least one attribute data to include at least a size request data determining a total number of the plurality of drives in the volume group;
automatically selecting a plurality of drives of the volume group based on a prioritized order of (i) an optimum tray distribution of the volume group (ii) a spindle speed of each drive in the volume group, wherein the spindle speed of the each drive in the volume group is no slower than a spindle speed of any one of unselected drives in the each of the drive trays, and (iii) a vertical alignment of the plurality of drives in the volume group; and
accessing each of the plurality of drives through a preferred controller I/O channel that is less utilized when a dispersion request data of the at least one attribute data is processed.
2. The method of claim 1 wherein the alignment of the plurality of drives in the volume group is achieved through selecting the plurality of drives based on slot locations of the plurality of drives.
3. The method of claim 2 further comprising constructing mirrored pairs with the plurality of drives based on a RAID conformance algorithm when the volume group is based on at least one of a RAID level 1 and a RAID level 10.
4. The method of claim 1 in a form of a non-transitory machine-readable medium embodying a set of instructions that, when executed by a machine, causes the machine to perform the method of claim 1 .
5. A system of a redundant array of independent drives (RAID), comprising:
a processor;
a user device to communicate an attribute data describing a volume group of the RAID;
a drive selection module to automatically select a plurality of drives of the volume group based on the attribute data using a baseline selection algorithm;
a tray distribution module of the baseline selection algorithm to iteratively select a single drive from each of drive trays of the volume group starting from a larger drive tray until a requested number of drives in the volume group is reached;
a spindle speed module of the baseline selection algorithm to select the plurality of drives with each of the plurality of drives having the spindle speed no slower than a spindle speed of any unselected drive in the each of drive tray;
an alignment module of the baseline selection algorithm to select the plurality of drives to achieve a vertical alignment of the plurality of drives;
wherein said modules of the baseline selection algorithm are prioritized in an order of the tray distribution module, the spindle speed module, and the alignment module;
a channel dispersion module of an extended selection algorithm to select preferred I/O channels of the volume group that are maximally dispersed among I/O channels of the RAID;
a RAID conformance algorithm module to automatically select the plurality of drives that are dispersed across an overall set of drive side I/O channels such that better read and write operations are achieved on the volume group; and
a best drive selection module to select the plurality of drives of the volume group based on a combination of the baseline selection algorithm, the extended selection algorithm, and the RAID conformance algorithm.
6. A method of a controller firmware of a redundant array of independent drives (RAID), comprising:
processing a plurality of attributes describing a volume group of the RAID;
constructing a list of tray groups with each member of the tray groups having at least one drive tray with at least one drive selectable to the volume group;
descendingly ordering the list of tray groups based on a number of the at least one drive;
automatically selecting a set of drives of the volume group based on a prioritized order of:
(i) an optimum tray distribution to evenly distribute the set of drives across a maximum number of drive trays; ,
(ii) a spindle speed such that a drive with a higher spindle speed in a particular tray is selected first before a drive with a lower spindle speed in the particular tray, and
(iii) a vertical alignment of the set of drives in the volume group based on slot locations of the set of drives.
7. The method of claim 6 wherein one of the plurality of attributes is a number of drives forming the volume group.
8. The method of claim 7 wherein the plurality of attributes to include a dispersion request data which triggers a selection of a particular drive with a less utilized preferred I/O channel.
9. A method of a redundant array of independent drives (RAID), comprising:
processing at least one attribute data describing a volume group of the RAID;
automatically selecting a plurality of drives of the volume group based on a prioritized order of:
a) an optimum tray distribution of the volume group,
b) a spindle speed of each drive in the volume group; and
c) a preferred controller I/O channel that is less utilized when a dispersion request data of the at least one attribute data is processed by accessing each of the plurality of drives.
10. The method of claim 9 , further comprising
d) aligning the plurality of drives in the volume group, wherein the at least one attribute data to include at least a size request data determining a total number of the plurality of drives in the volume group.
11. The method of claim 10 , wherein the at least one attribute data to include at least a size request data determining a total number of the plurality of drives in the volume group.
12. The method of claim 10 , wherein the spindle speed of the each drive in the volume group is no slower than a spindle speed of any one of unselected drives in the each of the drive trays.
13. The method of claim 10 further comprising constructing mirrored pairs with the plurality of drives based on a RAID conformance algorithm when the volume group is based on at least one of a RAID level 1 and a RAID level 10.
14. The method of claim 9 wherein the alignment of the plurality of drives in the volume group is achieved through selecting the plurality of drives based on slot locations of the plurality of drives.
15. The method of claim 14 , wherein the optimum tray distribution is obtained through selecting the plurality of drives such that the plurality of drives are substantially evenly distributed across a maximum number drive trays with each drive tray holding at least one selectable drive. drives vertically.