CN103345925A - Shingle-written magnetic recording device with hybrid E-regions - Google Patents
Shingle-written magnetic recording device with hybrid E-regions Download PDFInfo
- Publication number
- CN103345925A CN103345925A CN2013101314938A CN201310131493A CN103345925A CN 103345925 A CN103345925 A CN 103345925A CN 2013101314938 A CN2013101314938 A CN 2013101314938A CN 201310131493 A CN201310131493 A CN 201310131493A CN 103345925 A CN103345925 A CN 103345925A
- Authority
- CN
- China
- Prior art keywords
- district
- group
- exception record
- write
- order
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11B—INFORMATION STORAGE BASED ON RELATIVE MOVEMENT BETWEEN RECORD CARRIER AND TRANSDUCER
- G11B5/00—Recording by magnetisation or demagnetisation of a record carrier; Reproducing by magnetic means; Record carriers therefor
- G11B5/012—Recording on, or reproducing or erasing from, magnetic disks
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11B—INFORMATION STORAGE BASED ON RELATIVE MOVEMENT BETWEEN RECORD CARRIER AND TRANSDUCER
- G11B5/00—Recording by magnetisation or demagnetisation of a record carrier; Reproducing by magnetic means; Record carriers therefor
- G11B5/127—Structure or manufacture of heads, e.g. inductive
- G11B5/31—Structure or manufacture of heads, e.g. inductive using thin films
Landscapes
- Engineering & Computer Science (AREA)
- Manufacturing & Machinery (AREA)
- Signal Processing For Digital Recording And Reproducing (AREA)
- Digital Magnetic Recording (AREA)
Abstract
The invention relates to an SMR disk drive with hybrid E-regions, including a nonvolatile solid state memory E-region in addition to magnetic media E-regions on the disk. The memory E-region can be used in operations that can be referred as destaging and/or restaging to sequentialize sets of exception records to reduce the time and energy spent in executing seeks in the disk E-region. The ratio of the size of the solid state memory E-region to the total E-region capacity on the disks can be optimized for selected applications according to the invention using tradeoffs between performance and cost. For example, an embodiment with a memory E-region size that is 10% of the total disk E-region capacity achieves substantial performance enhancement over a disk-only E-region implementation and also results in smaller costs than can be required in a NAND-only memory E-region.
Description
Technical field
The present invention relates to the field of data storage device framework, relate in particular to and watt write magnetic recording (SMR) device.
Background technology
Conventional disc drive with magnetic medium is organized data with the concentric magnetic track that is spaced apart from each other.Watt writing concept is a kind of perpendicular magnetic recording form, and is described to a kind of mode that improves the surface density of magnetic recording.Watt writing in magnetic recording (SMR) medium, the district of adjacent track (band) is written in order to cover one or more magnetic tracks of before having write.With traditional can to carry out the separated magnetic track of write operation with random order different, watt magnetic track must carry out write operation in order.The magnetic track of panel surface is organized as a plurality of watts of folded districts (being also referred to as the I district).The I district watt write direction can be from interior diameter (ID) to overall diameter (OD) or from OD to ID.On same surface, also can adopt two endless belt (zone) to carry out watt folded (shingle) from both direction at dish in the mode that general mid diameter point joins.The quantity of watt magnetic track that stacks is a watt important parameter of writing in district.In case carry out write operation with watt stack structure, (in place) upgrades single magnetic track on the spot, because can rewrite and damage the data in the overlapping magnetic track like this.Therefore, from user's viewpoint, watt write data track and think similar daily record (log) that is used for splicing (oppend-only) sometimes.In order to improve the performance of SMR driver, the part of magnetic medium is assigned with as one or more so-called " exceptions area " (E district), and it is used as the outer level zone (staging area) for the data that finally are written to the I district.Sometimes the E district is considered to the E cache memory.Because the most of data during SMR drives are all expected to be stored in order in the I district, the record data in the current I of the not being stored in district can be considered to the unusual of order I district's storage.If receive the instruction that random sequence is write, they are stored in the E district of dish usually by order of receipt.
People's such as Sanvido (on June 21st, 2011) US Patent No. 7965465 has been described application caches technology of new record more in watt folded piece of the SMR disk storage that must carry out write operation in order.
Watt writing indirect address (address indirection) in the inside structure of memory storage is conducive to the complicacy that main frame is related with SMR and keeps apart.Traditional host file system uses LBA (Logical Block Addressing) (LBAs) to read or writing data blocks in order, does not consider the inner physical location of using (physical block address (PBA)) of memory storage.Recent decades, it is indirect that hard disk drive has possessed the LBA-PBA of certain level, its especially the bad sector on the permission dish be mapped at the good sector of for this reason reserving by replay.Typically, indirect address is often implemented in the controller part of driver architecture.Described controller converts the LBAs in the Host Command to the internal physical address or can finally be derived the intermediate address of physical address by it.
Shine upon traditional LBA-PBA for defective and do not need frequent changes.On the contrary, in the SMR driver, the physical block address (PBA) of LBA (Logical Block Addressing) (LBA) may frequent changes.Indirect system provides the dynamic translation layer between the current physical address on main frame LBAs and the medium.In the SMR system, because system dynamically determines the physical address of host data on medium of the LBA that will write, described LBA-PBA mapping can change according to each write operation.Because main frame LBA is updated, the data of same LBA will be written to different positions next time.In addition, driver mobile data between the I district of the E district of writing cache memory, dish of writing cache memory, dish of RAM and dish independently.No matter which driver is stored in to data, and the described LBAs of described data will remain unchanged.Background process as defragmentation is also independently carried out by described driver, thereby data sector is moved to another from a PBA, and LBA remains unchanged simultaneously.
Defragmentation is Essential Terms, thereby is used for describing the processing that fragment was eliminated or reduced to record in reorganization file or the Database Systems.In SMR I district, when record was updated or deletes, the quantity that is referred to as little free spaces invalid or " expired (stale) " data usually can increase.Defragmentation is handled physically, and thereby mobile this record makes it more continuous, generates bigger, more useful clear area simultaneously.DRAM is applied in during upgrading usually, effectively is categorized into suitable sequence because its permission will be recorded.In the SMR driver, the key factor in the whole performance that effective defragmentation is driver.
Summary of the invention
Embodiments of the invention comprise the SMR disk drive, and the magnetic medium E district on dish, it also possesses non-volatile solid state memory E district.The combination of solid-state memory and PanEQu is called as and mixes the E district.In the embodiment of the invention, solid-state memory E district can be used to be called the operation of degradation (destaging) and/or upgrading (restaging).Described storer E district can be used to make to organize exception record serializing (sequentialize) more, carries out time and the energy (engergy) that spends on the tracking thereby be reduced in PanEQu.The ratio workability energy of total E district capacity gone up by the size in solid-state memory E district and dish and the balance (tradeoff) between the cost is optimized selecteed application according to the present invention.For example, the size in storer E district is that 10% embodiment of total dish E district capacity can obtain significant performance with respect to the embodiment that has only PanEQu and strengthens, and compares and have only nand memory E district can obtain lower cost.
In one embodiment, writing of one group of input unusually at first is stored in solid-state memory E district, resequences before being downgraded to PanEQu then.The use in solid-state memory E district allows to resequence, thereby has improved sequentiality, and has reduced tracking required when data recording is submitted to disk storage.
In another embodiment, for example, as in the defragmentation process, non-volatile solid state memory E district will be used to upgrade and before write needs on the dish and reorder and then be rewritten to data on the described dish.Degradation and upgrading embodiment can use together.
Description of drawings
Figure 1 shows that the diagram that possesses the SMR data storage device that mixes the E district according to the embodiment of the invention.
Figure 2 shows that expression is for the chart that continues heavy read rate (the tracking number of per second) on the relative transverse axis of IOP on the Z-axis of 4 drivers in the mixing E district that possesses 1GB, 2GB, 4GB and 8GB nand flash memory of the SMR driver that has only PanEQu and the embodiment of the invention.
Figure 3 shows that the chart of the E district size that the expression PanEQu of having only used according to the invention and NAND mixing E district coupling PMR performance is demoted required.
Figure 4 shows that expression is for the histogram of the big or small shared user's percent of volume in dish E district of a plurality of equipment that possess the stressed cache size of multiple NAND E district size and DRAM and transition length.
Figure 5 shows that expression conventional P MR driver, have only among dish E district SMR driver and the present invention the sequence of MB/ second in random writing district in the mixing E district driver to read the chart of performance.
Figure 6 shows that and carry out the diagram that data block reorders during demoting from nand flash memory E district to PanEQu according to the embodiment of the invention.
Thereby Figure 7 shows that according to the embodiment of the invention uses nand flash memory E district will be convenient to the diagram of reordering of data block from the data block upgrading of PanEQu during the defragmentation of PanEQu.
Figure 8 shows that the diagram of during the defragmentation of PanEQu, getting back to PanEQu according to the embodiment of the invention from the nand flash memory E district degradation of the data block that reorders.
Embodiment
Figure 1 shows that the diagram of using the data storage device (DSD) 10 of the SMR that possesses a system electronic unit 21 with reference to one embodiment of the invention.Described system electronic unit 21 moves according to prior art except said, and the part of carrying out the prior art function in the described system is not shown.Described system electronic unit 21 can be the chip system of prior art, and it is integrated circuit, comprises the host interface on one chip, controller, servo function, microprocessor, firmware program etc. entirely.
Main frame/user 11 can be the computing machine of any kind, and can install by comprising any module communication via network with other.Term " user " and " main frame " replaceable use.A plurality of main frames also can be used prior art and communicate by letter with device.Film 12 is magnetic thin film coatings, be deposited on usually on the upper and lower surface of hard disk (not shown), and a device can have a plurality of dishes.Fig. 1 shows the sectional view of film 12.In plan view, each district is a plurality of concentric round belts.Magnetic thin film is formatted in order to be applied to the SMR framework, and in the present embodiment, it comprises dish E district 16, I district 13 (also can be described as I magnetic track district), writes cache memory area 14 (also can be described as twice and write cache memory area) and boot section or be with 15.Although have only one among Fig. 1,, a device can have a plurality of E district 16 in each panel surface, and a plurality of panel surface are arranged usually.Non-volatile solid state memory E district 19 is commonly called nand flash memory E district or NAND E district.According to the present invention, have only dish E area scheme relatively, described nand flash memory E district 16 is used for strengthening the property and saving energy.In addition, become remarkable in having only the required size in E district of NAND by the size design with NAND E district 16, can reduce cost.
Problem to be solved by this invention can coarsely analogize to postal delivery.Letter has random sequence, need be according to the specific destination that increases, and for example, country, area, state, Zip code, Mail Path, postal delivery are classified in proper order.The postman do not consider the address and at random backward or forward delivering mail be unusual poor efficiency.
Therefore, in this analogy, as mail, receive according at random order in essence and write unusually, still, according to the order of coming they are write dish and go up to cause that described unusual needs will be resequenced from now on the time and spend a large amount of time and resource at tracking.Before being write the I magnetic track in I district effectively, describedly need organize according to different modes unusually.High-grade the presorting of carrying out according to postcode can be compared to the tissue abnormalities according to target I district.Other grades of classifying are unusually comprised the LBA sequence.To classify unusually according to the present invention, and because the number of times of tracking reduces greatly, make the write operation that continues be subjected to the restriction of seeking speed hardly.
Except following description, general degradation and upgrading algorithm all can be used for managing in NAND E district and coil between the E district back and forth data transmission.In following description, the non-volatile solid state memory that is used as E district 19 is commonly referred to nand flash memory, because this is based on the preferred write buffer of present technology.Yet, the nonvolatile solid state memory buffer of any kind, for example phase transition storage, NOR flash memory and/or MRAM are in the present invention available.The application of term " NAND " is in order to simplify, but not means the restriction embodiments of the invention.
Use the degradation that mixes the E district
Use nand flash memory E district 19 to make one group when being downgraded to dish E district 16, can be re-ordered into sequence generally unusually as write buffer.Preferably, accumulation one big group exception record is classified before writing PanEQu.Therefore, degradation triggers when has expired in flash memory E district 19 usually, thereby maximizes the unusual quantity in each I district when degradation.Reduced the number of the tracking (read again and get) when carrying out defragmentation like this.
Figure 6 shows that according to the embodiment of the invention and during demoting to dish E district 16 from nand flash memory E district 19, carry out the diagram that the abnormal data piece reorders.During this period, main frame can be in order, inverted order, other nonrandom order or basic random sequence send write record.When with the random sequence receiving record, prior art the invention provides maximum gain relatively.In example shown in Figure 6, suppose to receive the LBA piece on the left side with original random sequence from main frame/user, and place it in nand flash memory E district.Although can in nand flash memory E district exception record be classified in order, the unusual physical sequential of LBA piece in flash memory E district is not crucial, because the movement from the entrance to the entrance does not need machinery, time consuming tracking.Key is that LBA piece group is downgraded to PanEQu according to the sequence order that improves.This means and unprocessedly among the SMR that dish E district 16 do not comprise as prior art write at random.One group of main frame is write initially reordering of (exception record) has at random increased the single unusual effectively stressed efficient of getting.The quantity of tracking has reduced, but this is independent of stressed speed.
These " skip order " are read in burst based on upgrading/defragmentation.The invention enables the quantity of the tracking at random in the defragmentation process in dish E district 16 to reduce greatly.Before one group of exception record being write dish E district 16 or writing fashionable reordering and to be realized by system electronic, and obtained by the software program of microprocessor execution.
Owing in having only dish E area scheme, need more to read at random, make it suffer a) the short tracking structure of difference; B) higher branch type load; C) the SRAM deficiency of large-scale internal queues; And d) better power consumption.Opposite, during upgrading/defragmentation, reading again when getting, the present invention can obtain tracking at random still less, can improve thereby use the above-mentioned variety of issue that has only dish E district designing institute to possess, and comprises and saves enough power consumptions.Although DRAM can substitute nand flash memory, it is more useful than DRAM to use NAND.NAND is than general cheap ten times of the every MB of DRAM, and different with DRAM, and NAND is non-volatile.
It is more cheap than the 100% application nand flash memory E alternative PanEQu in district to mix the design of E district.The size in NAND E district is fully less than total dish E district size among the present invention.For example, shown in following data, NANDE district size accounts for 10% of total dish E region memorizer size can improve performance.Use 10% NAND E district size and also the cost that increases can be reduced to 10% with respect to a design in NAND E district.Performance estimation (seeing Table 1) is illustrated in to mix in the design of E district uses 2GB NAND can obtain 498IOPs for the system of degradation, and its system performance with a NAND E district (421IOPs) that possesses 32GB is suitable.Estimation hypothesis 2TB user capacity in the table 1,4k RW IOPs.IOPs represents the I/O operation of per second.
| NAND capacity (GB) | The number percent that surpasses supply | NAND E district IOP | The NAND IOP that demotes |
| 2 | 0.11％ | 30 | 498 |
| 4 | 0.21％ | 59 | 568 |
| 8 | 0.43％ | 116 | 626 |
| 32 | 1.72％ | 421 | 783 |
| 64 | 3.44％ | 747 | 954 |
Table 1
Current perpendicular magnetic recording (PMR) Driving technique will write the position that is positioned at random at random, thus the more delay that before writing, needs more tracking at every turn and cause.The SMR driver in application mix E district possesses random writing IOPs faster, because do not need tracking at random during user data writes.New user data is write the I magnetic track in I district in order, perhaps writes the PanEQu in the SMR driving.
Figure 2 shows that expression uses the SMR driver have only PanEQu and possess the chart that weighs read rate (the tracking number of per second) on the Z-axis of 4 drivers in the mixing E district that uses 1GB, 2GB, 4GB and 8GB nand flash memory on the lasting relative transverse axis of IOPs.4 are mixed among the E district embodiment each all than having only PanEQu to obtain higher IOPs, and weight read rate scope is 200-1000 tracking number of per second.Although bigger nand flash memory can obtain more performance, however the flash memory of double sized than before single size can only obtain less lifting.
Unusual orderization has realized almost completely not being subjected to the influence of seeking speed and the write operation that obtains continuing among the present invention.NAND E district degradation (" NED ") has obtained and a performance that the design of NAND E district is equal, and has used the NAND of 80%-90% less.Can obtain identical performance gain if use the alternative NAND of the DRAM E district buffer memory of equal size, NAND E district is better than DRAM.
Chart among Fig. 3 will be used to have only among PanEQu and the present invention and use the long pass of demoting in NAND mixing E district and pass the block size performance and done contrast.Transverse axis is the transmission length of 1k unit.This estimates hypothesis PMR performance rate, command queue's degree of depth (QD)=32, and enable write cache memory (WCE)=1, for example WCE opens.Vertical axis represents disc E district size takies the number percent of family capacity.Size as fruit tray E district can be expanded, and heavy read rate can have a strong impact in order to satisfy the E district amount of space of PMR equivalent performance class requirement.The expansion of E district makes stressed going bail for hold dominant advantage.A given unusual heavy read rate, E district size the doubling according to each transmission length of requirement.Keeping the unusual average of every magnetic track like this is constant.
3 line representatives above in the chart have only the heavy read rate of per second 200,500 and 1000 tracking numbers in the dish E district.Following line representative mixing E of the present invention district.Suppose 1k transmission length, 200 heavy read rates, the invention enables satisfy the required dish E district size of PMR performance only account for have only dish E district big or small 10%.The advantage in mixing E district still can be not poorer than having only PanEQu along with the increase of transmission length descends to some extent.Possess or wish to obtain the driver of low heavy read rate, can use NAND degradation (" NED ") and reduce dish E district demand, can distribute more drive capacity to be used for storage of subscriber data to the I district like this.(note: other driver relatively, can obtain lower heavy read rate by the design balance, for example, save space, reduction tracking power, perhaps reduce cost.For example can select coil littler, that power consumption is lower, so reliably movement actuator.)
Use and mix the upgrading of E district
Except initial reordering before degradation as mentioned above, nand flash memory E district 19 also can be used to upgrading (for example, the defragmentation of I district or PanEQu).Figure 7 shows that thereby using nand flash memory E district 19 according to one embodiment of the invention during the defragmentation of PanEQu will promote the diagram of reordering of data block from the data block upgrading in dish E district 16.Figure 8 shows that according to one embodiment of the invention during the defragmentation of PanEQu from the nand flash memory E district of the data block that reorders to the diagram of PanEQu degradation.Can be by finishing upgrading by the system electronic of microprocessor software program for execution.As if when defragmentation dish E district, application flash memory E district from described dish E district reading and recording, like this can reduce seek time according to physical sequential.In case flash memory E district put in record, can easily use as the optional standard of LBA order and so on to come it is classified.
Using the benefit of upgrading in NAND E district comprises:
Reduce dram requirement (DRAM is than expensive ten times of nand flash memories), for example drop to 16MB to the DRAM of 64MB.The DRAM cache size is very important in dish E district size requirement.
Upgrading is the unusual additional opportunities that reorders before the I magnetic track re-writes.
Post in Fig. 4 chart is depicted as the number percent that the device mid-game E district size that possesses different parameters takies the family capacity.3 posts of every group are corresponding with the optional NAND E district that illustrates under post and the stressed cache size of getting of DRAM.For example, the NAND=2GB of the first group of post in the left side, DRAM=32MB.In every group of post, from left to right, described post corresponding 128k, 64k and 32k transmission length.The result shows that the DRAM cache size is very important in dish E district size requirement.Therefore, it is useful will re-writing NAND E district unusually before defragmentation.For example, by use NAND E district substitute DRAM upgrade can reduction DRAM size.
Under the situation of the NAND that can be used for demoting, it is useful will preserving back unusually among the NAND before the defragmentation of upgrading to NAND at PanEQu.Will follow a large amount of DRAM to use with the merging of I magnetic track takes place.For example, if the size in E district is ten times of NAND degradation buffer memory, before defragmentation, need average 10 times tracking number to recover all from coiling the unusual of E district I magnetic track.A plurality of unusual I magnetic tracks are read again in each tracking and are got, and have therefore increased the buffer memory demand, have reduced each I magnetic track and have read the cost of getting again.
Chart among Fig. 5 shows conventional P MR driver, have only among dish E district SMR driver and the present invention the sequential read performance of every MB/ second in random writing district in the mixing E district driver.Transverse axis is represented transmission length.The result shows to have only dish E district SMR driver relatively, and NAND E district to have reduced in order satisfying in the main frame district that comprises the random writing data and to read the desired quantity that reads at random in proper order among the present invention.
Ti Huanshishili ﹠amp; Optional attribute
Some optional attributes or the optimization that comprises among the various embodiment of the present invention is discussed below.An option is placed for the OD that will coil E district 16 adjacent discs, thereby uses hard disk at the higher revolution bit number at OD place.Near the OD magnetic track is longer, and have general 2 times to the data of the revolution of exemplary disc ID.
Another option is to select total PanEQu storage to account for 3% of total disk storage, and 30GB for example is in 1TB SMR driver.Because dish E district magnetic track is that cost is obtained to pay I district magnetic track, dish E district size is a benefit of the present invention less than its size in other modes.If use bigger NAND memory capacity, so Dui Ying bigger E district capacity can cause increasing progressively benefit and reduces to some extent.(as shown in Figure 2)
As discussed above, the storage of total NAND E district is decided to be total PanEQu storage general 10% or total disk storage 0.3%, being one can obtain the tempting selection that enough performances improve on the reasonable basis of cost.If total PanEQu storage is decided to be the general 3% of total disk storage, and NAND E district be total PanEQu store 10%, NAND E district accounts for 0.3% of total disk storage so.
The option of another consideration is that DRAM can be as buffer memory that will write operation before writing to NAND and/or dish; DRAM and SRAM are used by hard disk controller, thereby help reorder, mobile user data.
The sequence number of submitting on May 23rd, 2011 that the present invention relates to be commonly assigned is 61/489174 temporary patent application, and requires the right of priority of this provisional application according to 35U.S.C.119 (e).
This by reference in conjunction be commonly assigned on July 18th, 2011 submit to be numbered 13/135953 patented claim, this application has been described the SMR driver embodiment that writes twice cache memory area that has that mentions in the present invention.
Claims (13)
1. one kind of watt of magnetic recording disk drive comprises:
Dish with magnetic thin film coating, data are stored in the concentric circles magnetic track;
The I district, it comprises first subclass of magnetic track, covers previous magnetic track with track portion ground and sequentially carries out watt writing;
PanEQu, it comprises second subclass of magnetic track, is used for keeping exception record;
Storer E district, it comprises the exception record that is stored in the non-volatile solid state memory; And
System electronics, its first order when reading from described PanEQu upgrades to described storer E district with one group of exception record from described PanEQu, then described one group of exception record is write described PanEQu according to being different from described first in proper order the sequence order.
2. disk drive as claimed in claim 1 also comprises a plurality of PanEQu, and it possesses total dish E district storage size, and wherein, the size in described storer E district is significantly less than described total dish E district storage size.
3. disk drive as claimed in claim 2, the size in wherein said storer E district accounts for 10% of total dish E district storage size greatly.
4. disk drive as claimed in claim 2, wherein said sequence order make the seek time of writing described one group of exception record and needing be less than according to the described one group of needed seek time of exception record of first sequential write when reading.
5. disk drive as claimed in claim 1, wherein said sequence order causes that described one group of exception record is write the needed seek time in I district and is less than when reading according to the described one group of needed seek time of exception record of described first sequential write.
6. disk drive as claimed in claim 1, wherein said system electronic unit will be stored in according to second group of exception record that the reception order receives in write order in the described storer E district, and according to the sequence order that is different from described reception order described second group of exception record be write described PanEQu subsequently.
One kind the operation watt magnetic recording disk drive method, comprising:
One group of exception record is read storer E district non-volatile solid state memory from PanEQu, and described one group of exception record reads in proper order according to corresponding with physical sequential in described dish E district first; And
Described one group of exception record is write PanEQu according to the sequence order that is different from first order, and described sequence order is according to the classified order corresponding with the selected feature of described one group of exception record.
8. method as claimed in claim 6, wherein said sequence order cause that described one group of exception record is write the required seek time in I district and are less than the described one group of seek time that exception record is required of described first sequential write when reading.
9. the method for operation board driver as claimed in claim 8, wherein said disk drive also comprises a plurality of PanEQu with total dish E district storage size, and the size in wherein said storer E district is significantly less than described total dish E district storage size.
10. method as claimed in claim 9, the size in wherein said storer E district account for 10% of described total dish E district storage size greatly.
11. as the method for claim 10, wherein said sequence order causes that described one group of exception record is write the required seek time in I district and is less than the described one group of seek time that exception record is required of described first sequential write when reading.
12. method as claimed in claim 7 also comprises: before according to sequence order described one group of exception record being write described PanEQu, described one group of exception record is categorized into the sequence order in the storer E district.
13. method as claimed in claim 7 also comprises:
Store second group of exception record into described storer E district, described second group of exception record receives in write order according to random sequence; And
According to second sequence order that is different from described random sequence described second group of exception record write described PanEQu, described second sequence order is the classified order for the selected feature of described second group of exception record.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US13/411,503 US8681443B2 (en) | 2011-05-23 | 2012-03-02 | Shingle-written magnetic recording (SMR) device with hybrid E-region |
| US13/411,503 | 2012-03-02 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103345925A true CN103345925A (en) | 2013-10-09 |
Family
ID=48985888
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2013101314938A Pending CN103345925A (en) | 2012-03-02 | 2013-03-04 | Shingle-written magnetic recording device with hybrid E-regions |
Country Status (3)
| Country | Link |
|---|---|
| KR (1) | KR101526042B1 (en) |
| CN (1) | CN103345925A (en) |
| DE (1) | DE102013003545A1 (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106548789A (en) * | 2015-09-17 | 2017-03-29 | 伊姆西公司 | Method and apparatus for operating stacked tile type magnetic recording equipment |
| CN106663451A (en) * | 2014-06-27 | 2017-05-10 | 国际商业机器公司 | Hybrid surface format hard disk drive |
| CN108021513A (en) * | 2016-11-02 | 2018-05-11 | 杭州海康威视数字技术股份有限公司 | A kind of date storage method and device |
| CN110275677A (en) * | 2019-05-22 | 2019-09-24 | 华为技术有限公司 | Hard disk form conversion method, device and storage equipment |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7965465B2 (en) | 2009-03-11 | 2011-06-21 | Hitachi Global Storage Technologies Netherlands, B.V. | Techniques for storing shingle blocks in a cache memory using a data storage device |
| US8271801B2 (en) * | 2009-11-19 | 2012-09-18 | Hitachi Global Storage Technologies Netherlands B.V. | Implementing data confidentiality and integrity of shingled written data |
-
2013
- 2013-02-28 KR KR1020130021696A patent/KR101526042B1/en not_active IP Right Cessation
- 2013-03-01 DE DE201310003545 patent/DE102013003545A1/en not_active Withdrawn
- 2013-03-04 CN CN2013101314938A patent/CN103345925A/en active Pending
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106663451A (en) * | 2014-06-27 | 2017-05-10 | 国际商业机器公司 | Hybrid surface format hard disk drive |
| CN106663451B (en) * | 2014-06-27 | 2019-07-09 | 国际商业机器公司 | Blending surface format hard disk drive |
| CN106548789A (en) * | 2015-09-17 | 2017-03-29 | 伊姆西公司 | Method and apparatus for operating stacked tile type magnetic recording equipment |
| CN106548789B (en) * | 2015-09-17 | 2019-05-17 | 伊姆西公司 | Method and apparatus for operating stacked tile type magnetic recording equipment |
| CN108021513A (en) * | 2016-11-02 | 2018-05-11 | 杭州海康威视数字技术股份有限公司 | A kind of date storage method and device |
| CN108021513B (en) * | 2016-11-02 | 2021-09-10 | 杭州海康威视数字技术股份有限公司 | Data storage method and device |
| CN110275677A (en) * | 2019-05-22 | 2019-09-24 | 华为技术有限公司 | Hard disk form conversion method, device and storage equipment |
| CN110275677B (en) * | 2019-05-22 | 2022-04-12 | 华为技术有限公司 | Hard disk format conversion method and device and storage equipment |
| US11682428B2 (en) | 2019-05-22 | 2023-06-20 | Xfusion Digital Technologies, Co., Ltd. | Hard disk format conversion method and apparatus, and storage device |
Also Published As
| Publication number | Publication date |
|---|---|
| KR20130100727A (en) | 2013-09-11 |
| DE102013003545A1 (en) | 2013-09-05 |
| KR101526042B1 (en) | 2015-06-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103295594A (en) | Shingle-written magnetic recording (SMR) device with hybrid E-region | |
| US20140122774A1 (en) | Method for Managing Data of Solid State Storage with Data Attributes | |
| US20100312950A1 (en) | Electronic storage device and control method thereof | |
| CN103135945B (en) | Multi-channel dynamic read-write dispatching method used in solid state drive (SSD) | |
| US20130254458A1 (en) | Single-level cell and multi-level cell hybrid solid state drive | |
| US9460751B2 (en) | Binding shingled recording bands | |
| CN106708751A (en) | Storage device including multi-partitions for multimode operations, and operation method thereof | |
| CN103345925A (en) | Shingle-written magnetic recording device with hybrid E-regions | |
| CN101727395A (en) | Flash memory device and management system and method thereof | |
| CN102043593B (en) | Region-based management method for external cache of disk | |
| US10802739B1 (en) | Data storage device configuration for accessing data in physical realms | |
| CN101458932A (en) | Method and apparatus for writing data with sequential access in a disk drive | |
| CN102999428A (en) | Four-stage addressing method for tile recording disk | |
| US20030149838A1 (en) | Data transfer performance through resource allocation | |
| US9378769B2 (en) | System and method for storing data with host configuration of storage media | |
| CN106527987A (en) | Non-DRAM SSD master control reliability improving system and method | |
| CN102362263A (en) | SSD controller, and method for operating an SSD controller | |
| US9658964B2 (en) | Tiered data storage system | |
| US7523255B2 (en) | Method and apparatus for efficient storage and retrieval of multiple content streams | |
| CN112000296B (en) | Performance optimization system in full flash memory array | |
| US20190220215A1 (en) | Data storage device and data storage control method | |
| CN106055266A (en) | Disk device and controlling method of disk device | |
| US20140258668A1 (en) | Systems and Methods for Managing Storage Space in Hybrid Data Storage Systems | |
| KR101577495B1 (en) | Shingle-written magnetic recording(smr) device with hybrid e-region | |
| CN116795732A (en) | Memory circuit, memory and memory address allocation method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20131009 |
|
| WD01 | Invention patent application deemed withdrawn after publication |