The patent's inventors are Sato, Katsuya (Odawara, JP); Inoue, Shintaro (Odawara, JP);
This patent was filed on
From the background information supplied by the inventors, news correspondents obtained the following quote: "Storage systems, due to the need to handle large volumes of data, have become larger in scale and more complex in recent years. For this reason, there is demand for storage systems that can be managed easily and used efficiently. One prior art that meets this demand is thin provisioning technology. In this technology, the storage system constructs a virtual logical volume (called a 'virtual volume' hereinafter) that does not comprise a physical storage area. The storage system then provides the virtual volume to a host computer. The host computer carries out a data write request to a virtual page that makes up this virtual volume. The storage system allocates a pool page from a pool configured from multiple pool pages to the virtual page targeted by the data write request. The pool comprises either one or multiple physical storage areas (referred to as a 'disk volume' hereinafter). Each disk volume is partitioned into two or more pool pages. The disk volume comprising the pool may be a substantial logical volume configured based on one or more storage media (for example, a RAID group configured using multiple storage media), or a virtual logical volume to which is associated a logical volume of a storage system external to the storage system.
"In accordance with this technology, the storage system-provided volume capacity can be efficiently used, and the volume capacity can be appropriately designed.
"The disk volume is generally configured from storage media of the same performance. However, the frequency with which respective virtual pages are accessed is normally not the same, and frequently accessed virtual pages are mixed together with infrequently accessed virtual pages. For this reason, for example, in a case where the disk volume is configured using only a high-performance storage medium, an excess of performance will occur with respect to a virtual page with a low access frequency. By contrast, in a case where a pool is configured using only a low-performance disk volume, a shortage of performance will occur with respect to a virtual page with a high access frequency.
"One method for solving this problem is disclosed in Patent Literature 1. According to Patent Literature 1, a single pool is configured using multiple disk volumes having different performance (typically, access performance), and based on the access frequency of each virtual page, a pool page allocated to a virtual page and data stored in this pool page is reallocated to a pool page of a disk volume comprising the ideal performance. The performance of the disk volume depends on the performance (typically, access performance) of the storage medium constituting the basis for this disk volume."
Supplementing the background information on this patent, VerticalNews reporters also obtained the inventors' summary information for this patent: "Technical Problem
"In Patent Literature 1, the reallocation of the pool page allocated to the virtual page is carried out subsequent to a change having been detected in the access frequency of the virtual page. The reallocation process takes time. Therefore, when a reallocation is executed after detecting a change in the access frequency of the virtual page as in the Cited Literature 1, a situation occurs in which the allocation of the pool page to the virtual page is not optimal.
"Accordingly, an object of the present invention is to lower the probability of a situation occurring in which the pool page allocation with respect to the virtual page is not optimal.
"Solution to Problem
"The storage system comprises a virtual volume, which is a virtual logical volume comprising multiple virtual page groups, a pool, which is a storage area comprising multiple pool page groups having different access performance, and a controller, which comprises a corresponding relationship between the virtual page and the pool page, and accesses the pool page corresponding to the virtual page based on accesses to the virtual page from other apparatuses. The controller, based on an access load (for example, frequency of accesses to the pool page) trend with respect to each virtual page in each time slot, executes processing for deciding an ideal pool page allocation destination for each virtual page in each time slot, and reallocating a virtual page to the ideal pool page at a time prior to a start time of each time slot so that the ideal pool page can be allocated to each virtual page by the start time of each time slot.
"In accordance with this, an ideal pool page is allocated to a virtual page in each time slot. Consequently, the virtual volume as a whole is able to demonstrate high performance in all of the time slots.
"In the preferred embodiment, when a cumulative access load for a virtual page from a start time of a certain time slot to a current time becomes equal to or larger than a cumulative access load for the virtual page in a certain time slot in the past that includes the current time, the current time is a preliminary reallocation start time, and a time which goes back by a minimum period of time required to complete the reallocation process from the start time of the certain time slot, is the reallocation start time. Then, in a case where the preliminary reallocation start time is earlier than the reallocation start time, the reallocation process starts from the preliminary reallocation start time, and in a case where the preliminary reallocation start time is either identical to or later than the reallocation start time, the reallocation process starts from the reallocation start time. For example, a 'certain time slot in the past that includes the current time' may be the same time slot as the time slot comprising 'from a start time of a certain time slot to a current time'.
"This makes it possible to carry out reallocation processing over a longer period of time. Consequently, the access load on the pool page generated at reallocation processing time can be temporally leveled.
"Furthermore, the storage system may be a single configuration or may be a multiple configuration."
For the URL and additional information on this patent, see: Sato, Katsuya; Inoue, Shintaro;
Keywords for this news article include: Technology,
Our reports deliver fact-based news of research and discoveries from around the world. Copyright 2014, NewsRx LLC
Most Popular Stories
- National Retail Federation Reduces Sales Forecast
- Desktop, Laptop Setups Still King
- Execs Help Entrepreneurs, Get Chevy Volts
- Demand for Fair Trade Brings Big Opportunity, Clear Conscience
- Shania Twain's Vegas residency ending after 110 shows
- Zillow in Reported $2B Bid for Real Estate Rival Trulia
- Google Confirms $1B Acquisition of Twitch.tv
- Sporty Ford Fiesta Fires on All 3 Cylinders
- Honda' s Accord Plug-in Hybrid Is a Fuel Miser
- Amazon Hiring on Calif.'s Central Coast