System and method for allocating resources in a mixed SSD and HDD storage environment

Patent metadata

Field	Value
Publication number	US-8996450-B1
Application number	US-201213731057-A
Country	US
Kind code	B1
Filing date	Dec 30, 2012
Priority date	Dec 31, 2011
Publication date	Mar 31, 2015
Grant date	Mar 31, 2015

Abstract

A system and method for allocating storage devices within a database system including fast solid state storage devices (SSDs) and traditional hard drive storage device (HDDs). The method repetitively collects temperature and queue depth data for each storage device, and determines from the collected data points a service time for each storage device. The method then selects, in response to an allocation request, a storage device having a minimum total wait time determined from the current queue depth and service time for each storage device.

First claim

What is claimed is:

1. A computer-implemented method for allocating resources in an electronic data storage environment comprising multiple electronic data storage devices, the method comprising the steps of:
repetitively, by a computer, collecting data points comprising queue depths (L Q ) and temperatures (r i ) corresponding to each storage device; and for each storage device, calculating from said set of data points (r i , L Q ), a service time (T ser );
receiving, by said computer, a data allocation request including an estimated frequency of access for said data;
for each storage device, calculating, by said computer, a total wait time from said estimated frequency of access for said data, a current queue depth for said each storage device, and the service time for said each storage device;
selecting, by said computer, a storage device having a minimum total wait time; and
returning, by said computer, the selected storage device for data allocation.

2. The method for allocating resources in an electronic data storage environment comprising multiple electronic data storage devices according to claim 1 , wherein said multiple electronic data storage devices comprise solid state drives and hard disk drives.

3. The method for allocating resources in an electronic data storage environment comprising multiple electronic data storage devices according to claim 1 , wherein said steps of collecting data points and calculating service times (T ser ) is repeated at a fixed interval.

4. The method for allocating resources in an electronic data storage environment comprising multiple electronic data storage devices according to claim 1 , wherein said electronic data storage environment comprises a relational database system.

5. An electronic data storage system, comprising:
multiple electronic data storage devices; and
a processor for:
repetitively collecting data points comprising queue depths (L Q ) and temperatures (r i ) corresponding to each storage device; and for each storage device, calculating from said set of data points (r i , L Q ), a service time (T ser );
receiving a data allocation request including an estimated frequency of access for said data;
for each storage device, calculating a total wait time from said estimated frequency of access for said data, a current queue depth for said each storage device, and the service time for said each storage device;
selecting for allocation a storage device having a minimum total wait time; and
returning the selected storage device for data allocation.

6. The electronic data storage system according to claim 5 , wherein said multiple electronic data storage devices comprise solid state drives and hard disk drives.

7. The electronic data storage system according to claim 5 , wherein said processor collects data points and calculates service times (T ser ) at a fixed interval.

8. The electronic data storage system according to claim 5 , wherein said electronic data storage system comprises a relational database system.

Assignees

• Teradata Corp
• Teradata Us Inc

Inventors

• Rubio Paul Anthony

Classifications

• G06F3/0659

  Command handling arrangements, e.g. command buffers, queues, command scheduling · CPC title

• G06F17/30592 Primary

  Physics · mapped topic

• G06F16/283 Primary

  Multi-dimensional databases or data warehouses, e.g. MOLAP or ROLAP · CPC title

• G06F3/0611

  in relation to response time · CPC title

• G06F3/0616

  in relation to life time, e.g. increasing Mean Time Between Failures [MTBF] · CPC title