TWI687921B - Host, shingled magnetic recording hard disk drive and operation method thereof - Google Patents

Abstract

The disclosure provides a host, a shingled magnetic recording (SMR) hard disk drive (HDD) and an operating method thereof. The SMR HDD includes a main storage area, a media cache, and a hard disk controller. The main storage area includes a plurality of stacked first tracks. The media cache includes a plurality of second tracks that surround the main storage area, wherein data are written to the second tracks in response to a plurality of write requests from the host. The hard disk controller is configured to: monitor an usage of the media cache; in response to the usage of the media cache reaching a first threshold of a plurality of usage thresholds, send a first link packet to the host to request the host to add a first delay time corresponding to the first threshold between the write requests.

Description

Host, shingled magnetic recording hard disk and operation method thereof

The invention relates to a host, shingled magnetic recording (SMR) hard disk and its operating method, and in particular to a media cache based on the SMR hard disk and adding a delay between write requests Time host, SMR hard drive and its operation method.

At present, the mainstream magnetic technology of traditional hard disk drives (HDD) has evolved into SMR. Unlike super giant magnetoresistance (CMR) HDDs, SMR HDDs magnetically stack the magnetic tracks of the main store area like roof tiles to increase the magnetic density. Since the stacked magnetic tracks increase the complexity of writing data, the delay time when performing random writing will be so long that it will drag on the processing capacity. Therefore, the SMR HDD is still designed and arranged in the CMR way on the outer track of the disc, and this area is called the media cache. When the host connected to the SMR HDD performs random data writing to the SMR HDD, the host will first write the data into the media cache to maintain the performance during random writing.

After the random data is written to the media cache, the SMR HDD will perform background activities during idle time when no host command is received to move the data from the media cache to the main storage area where the data should be placed Magnetic track. However, when the media cache area is full and the host continuously issues random write requests, the SMR HDD needs to first move a part of the data from the media cache to the main storage area, and then write the new data to the media cache . Such repeated actions will make the SMR HDD processing time very slow.

In view of this, the present invention proposes a host, SMR hard disk and its operating method, which can be used to solve the above technical problems.

The invention provides a shingled magnetic recording hard disk, which is connected to a host. The shingled magnetic recording hard disk includes a main storage area, a media cache, and a hard disk controller. The main storage area includes a plurality of stacked first magnetic tracks. The media cache includes multiple second magnetic tracks surrounding the main storage area, which are written with multiple data in response to multiple write requests from the host. The hard disk controller is configured to: monitor a usage of the media cache; respond to the usage of the media cache reaching a first threshold among the multiple usage thresholds, and send a first link packet to the host To request the host to add a first delay time corresponding to the first threshold between the aforementioned write requests.

The invention provides an operation method of a shingled magnetic recording hard disk, wherein the shingled magnetic recording hard disk is connected to a host and includes a main storage area, a media cache and a hard disk controller. The main storage area includes A plurality of stacked first magnetic tracks, the media cache includes a plurality of second magnetic tracks surrounding the main storage area, which is written with multiple data in response to multiple write requests of the host. The method includes: the hard disk controller monitors a usage amount of the media cache; and, in response to the media cache usage reaching a first threshold value among the plurality of usage threshold values, the hard disk controller sends A first link packet is sent to the host to require the host to add a first delay time corresponding to the first threshold between the aforementioned write requests.

The present invention provides a host connected to a shingle-type magnetic recording hard disk, including: a processor configured to issue multiple write requests to the shingle-type magnetic recording hard disk to write multiple data to A media cache for a shingled magnetic recording hard disk; receiving a first link packet from the shingled magnetic recording hard disk, wherein the shingled magnetic recording hard disk reaches multiple usage in response to the usage of the media cache A first threshold value among the threshold values requires the host to add a first delay time between the aforementioned write requests through the first link packet; the first link packet is parsed and the corresponding Time plus the first delay time.

Based on the above, in the present invention, when the SMR HDD performs random writing, it can self-monitor the usage of the media cache, and require the host to add a delay time between write requests when the usage reaches the first threshold . In this way, it is possible to avoid the situation that the system service time of the SMR HDD is too long due to full media cache.

In order to make the above-mentioned features and advantages of the present invention more obvious and understandable, the embodiments are specifically described below in conjunction with the accompanying drawings for detailed description as follows.

Please refer to FIGS. 1A and 1B, wherein FIG. 1A is a top view of an SMR HDD according to an embodiment of the present invention, and FIG. 1B is a schematic diagram of a main storage area according to FIG. 1B. In this embodiment, the SMR HDD 100 includes a main storage area 102, a media cache 104, and a hard disk controller 106. The main storage area 102 includes a plurality of stacked first magnetic tracks 102a, as shown in FIG. 1B. It should be understood that although each of the first magnetic rails 102a is shown as a horizontally long strip in FIG. 1B, it is only used to help understand the stacking manner of the first magnetic rails 102a. In an actual structure, each first magnetic rail 102a should be ring-shaped, but the present invention may not be limited to this.

The media cache 104 includes a plurality of second magnetic tracks 104a surrounding the main storage area 102. The hard disk controller 106 can be controlled by the host 199 to control the reading head 106a and the writing head 106b to read/write data to the main storage area 102 and the media cache. In different embodiments, the host 199 is, for example, a chipset in a computer, a central processing unit, or other components that can control the SMR HDD 100, but it is not limited thereto.

In an embodiment, the media cache 104 may be written to multiple data in response to multiple write requests WR (which are, for example, random write requests) of the host 199. More specifically, when the hard disk controller 106 receives the write request WR from the host 199, it accordingly controls the write head 106b to write the data to the media cache 104, and when the write request WR is not received At the time, the data in the media cache 104 is moved to the main storage area 102.

In one embodiment, in order to avoid the situation that the data in the media cache 104 cannot be moved to the main storage area 102 in real time and increase the processing time of the SMR HDD 100, the present invention proposes the following methods to improve SMR The operating efficiency of the HDD 100.

Please refer to FIG. 2, which is a flowchart illustrating an operation method of the SMR HDD according to an embodiment of the invention. The method of this embodiment can be performed by the SMR HDD 100 of FIGS. 1A and 1B. The details of each step of FIG. 2 are described below with the components shown in FIGS. 1A and 1B.

In step S210, the hard disk controller 106 may monitor the usage of the media cache 104. In one embodiment, the hard disk controller 106 can execute self-monitoring analysis and reporting technology (SMART). Specifically, SMART is an automatic hard disk status detection and early warning system and specification, which can perform the execution of the hard disk hardware (such as head, disk, motor, circuit) through the detection instructions in the hard disk hardware Monitor and record. In SMART, in addition to the attribute values that have been included in the public specification, there is a vendor-specific register (Vendor Specific Register), which allows the vendor (Vendor) to define the information that needs to be recorded.

Therefore, in one embodiment, the hard disk controller 106 can execute SMART to periodically or irregularly monitor the usage of the media cache 104 and record the usage of the media cache 104 in SMART's vendor-specific register However, the present invention may not be limited to this. In this embodiment, the vendor-specific register used to record the usage of the media cache 104 may be redefined as a media cache usage register, which may record the current usage of the media cache 104. In one embodiment, the usage of the media cache 104 may be characterized as a percentage, for example. For example, if 50% of the area of the media cache 104 is currently used to store data, the usage of the media cache 104 can be 50%, and the hard disk controller 106 can temporarily use the media cache. The memory records 50%. In addition, if 70% of the area of the media cache 104 is currently used to store data, the usage of the media cache 104 can be 70%, and the hard disk controller 106 can be used in the media cache usage register. Record 70%, but the invention may not be limited to this.

After that, in step S220, in response to the usage of the media cache 104 reaching the first threshold among the multiple usage thresholds, the hard disk controller 106 may send a first link packet to the host 199, to require the host 199 to add a first delay time corresponding to the first threshold value between the aforementioned write requests WR.

In addition, in other embodiments, in response to the usage of the media cache 104 reaching a second threshold value among the aforementioned usage threshold values, the hard disk controller 106 may send a second link packet to the host 199 to The host 199 is required to replace the first delay time with the second delay time corresponding to the second threshold value. In other words, when the hard disk controller 106 determines that the usage of the media cache 104 has reached another threshold among the aforementioned usage thresholds, the hard disk controller 106 may issue a link corresponding to the other threshold The packet is sent to the host 199 to request the host 199 to adjust the delay time between the write requests WR. In an embodiment, if the other threshold value is less than the first threshold value, the hard disk controller 106 may request the host 199 to shorten the delay time between write requests WR. Conversely, if the other threshold value is greater than the first threshold value, the hard disk controller 106 may require the host 199 to increase the delay time between write requests WR, but the invention may not be limited to this.

表1 From the above, different usage thresholds correspond to different delay times and link packets. Moreover, in the embodiments of the present invention, a higher usage threshold may correspond to a longer delay time. For ease of description, the correspondence between the usage threshold value and the delay time is provided below, but it is only used as an example and is not intended to limit possible implementations of the present invention.

Table 1

Taking Table 1 as an example, when the usage of the media cache 104 reaches 60%, the hard disk controller 106 may send a link packet to the host 199 accordingly to request the host 199 to add between the aforementioned write requests WR The value is a delay time of 1 ms. In addition, when the usage of the media cache 104 reaches 70%, the hard disk controller 106 may send another link packet to the host 199 accordingly, requiring the host 199 to add a value of 5 between the aforementioned write requests WR Delay time in ms. It should be understood that when the usage of the media cache 104 reaches 50%, it means that the media cache 104 is in a relatively empty state, so the hard disk controller 106 can write between the write requests WR through the corresponding link packets The delay time is adjusted to 0 ms (ie, no delay) to allow the SMR HDD 100 to operate in a conventional manner, but the present invention is not limited to this.

In one embodiment, the hard disk controller 106 may send the first link packet to the host 199 based on a device initiated power management (DIPM) technology. Accordingly, the host 199 can also be programmed to have the ability to parse the first link packet. Moreover, in an embodiment, after the host 199 receives the link packet from the SMR HDD 100, it can also decide whether to add a delay time between the write requests WR according to its current situation. If yes, the host 199 can return an acknowledgement (ACK) message to the SMR HDD 100 accordingly, and add the delay time corresponding to the link packet between the write requests WR sent later. Conversely, if the host 199 has an instruction that needs to be processed in time, the host 199 may respond to a negative acknowledge (NAK) message to the SMR HDD 100.

In order to make the concept of the present invention clearer, the following uses FIG. 3 as an example for further explanation. Please refer to FIG. 3, which is a schematic diagram of an application scenario according to an embodiment of the present invention. In FIG. 3, it is assumed that the host 199 transmits multiple write requests WR to the SMR HDD 100, and these write requests WR will be written to the media cache 104 of the SMR HDD 100 as described previously. At the same time, the hard disk controller 106 will monitor and record the usage of the media cache 104.

In an embodiment, assuming that the usage of the media cache 104 reaches 60%, the hard disk controller 106 may issue a corresponding link packet P1 to the host 199 according to the content of Table 1, to request the host 199 to write a request The delay time DT1 corresponding to 60% (ie, 1 ms) is added between the WRs.

After the host 199 receives and parses the link packet P1, if the host 199 determines according to the current situation that it can add the delay time specified by the link packet P1 between the write request WR, it can reply the confirmation message ACK to the SMR HDD 100, And the delay time DT1 is added between the write requests WR sent afterwards. In this case, the hard disk controller 106 can move the data in the media cache 104 to the main storage area 102 within the delay time DT1.

In addition, assuming that the usage of the media cache 104 rises to 70%, the hard disk controller 106 can also send a corresponding link packet P2 to the host 199 according to the content of Table 1, to request the host 199 to write the request WR Time plus delay time DT2 corresponding to 70% (ie, 5 ms).

After the host 199 receives and parses the link packet P2, if the host 199 determines according to the current situation that it can add the delay time specified by the link packet P2 between the write request WR, it can reply the confirmation message ACK to the SMR HDD 100, And the delay time DT2 is added between the write requests WR sent afterwards. In this case, the hard disk controller 106 can move the data in the media cache 104 to the main storage area 102 within the delay time DT2.

In other embodiments, if the usage of the media cache 104 drops to 50% (or lower), the hard disk controller 106 may also issue a corresponding link packet (not shown) according to the contents of Table 1. The host 199 requires the host 199 to add a delay time corresponding to 50% (ie, 0 ms) between the write requests WR. After the host 199 receives and parses the link packet, if the host 199 determines according to the current situation that it can add the delay time specified by the link packet between the write request WR, it can reply the confirmation message ACK to the SMR HDD 100, and A delay time of 0 ms (ie, no delay) is added between the write requests WR sent thereafter.

It should be understood that although the host 199 in FIG. 3 adds a delay time between any two of the write requests WR, in other embodiments, the host 199 may also be adjusted to add only one delay for every N write requests WR Time (N is an integer greater than 1), not limited to the form shown in Figure 3.

In summary, in the present invention, when the SMR HDD performs random writing, it can self-monitor the usage of the media cache, and then determine whether to require the host to add a delay time between write requests. If the usage of the media cache reaches a certain usage threshold, the SMR HDD can send a corresponding link packet to notify the host. After the host responds with the confirmation message, the host can add the delay time specified by the link packet between the write requests sent thereafter. In this way, SMR HDD can move the data in the media cache to the main storage area in a timely manner, thereby avoiding the situation that the system service time is too long due to the full media cache.

Although the present invention has been disclosed as above with examples, it is not intended to limit the present invention. Any person with ordinary knowledge in the technical field can make some changes and modifications without departing from the spirit and scope of the present invention. The scope of protection of the present invention shall be subject to the scope defined in the appended patent application.

100: SMR HDD 102: Main storage area 102a: first track 104: Media cache 104a: second track 106: Hard disk controller 106a: read head 106b: write head 199: Host ACK: confirmation message DT1, DT2: delay time P1, P2: link packet S210, S220: steps WR: write request

FIG. 1A is a top view of an SMR HDD according to an embodiment of the invention. FIG. 1B is a schematic diagram of the main storage area shown in FIG. 1B. FIG. 2 is a flowchart of an SMR HDD operating method according to an embodiment of the invention. FIG. 3 is a schematic diagram of an application scenario according to an embodiment of the invention.

S210, S220: steps

Claims (13)

A shingled magnetic recording hard disk, which is connected to a host, includes: A main storage area, which includes a plurality of stacked first magnetic tracks; A media cache, which includes a plurality of second magnetic tracks surrounding the main storage area, which is written with multiple data in response to multiple write requests of the host; A hard disk controller, which is configured to: Monitor the usage of the media cache; In response to the usage of the media cache reaching a first threshold among the usage thresholds, a first link packet is sent to the host to request the host to add between the write requests Up corresponds to a first delay time of the first threshold value. The shingled magnetic recording hard disk as described in item 1 of the patent scope, wherein the hard disk controller monitors the usage of the media cache based on a self-monitoring, analysis and reporting technology, and caches the media The usage is recorded in a vendor-specific register of the self-monitoring, analysis and reporting technology. The shingled magnetic recording hard disk as described in item 1 of the patent application scope, wherein the hard disk controller is further configured to: In response to the usage of the media cache reaching a second threshold among the usage thresholds, a second link packet is sent to the host to request the host to replace the first delay time with the corresponding A second delay time at the second threshold. The shingled magnetic recording hard disk as described in item 1 of the patent application scope, wherein the usage thresholds correspond to a plurality of delay times, the delay times include the first delay time, and the first threshold value It is positively related to this first delay time. The shingled magnetic recording hard disk as described in item 1 of the patent application scope, wherein the hard disk controller sends the first link packet to the host based on a device-initiated power management technology. The shingled magnetic recording hard disk as described in item 1 of the patent scope, wherein the hard disk controller moves the data to the main storage area in response to the first delay time between the write requests . An operation method of a shingled magnetic recording hard disk, wherein the shingled magnetic recording hard disk is connected to a host and includes a main storage area, a media cache and a hard disk controller, the main storage area includes a stack A plurality of first magnetic tracks, the media cache includes a plurality of second magnetic tracks surrounding the main storage area, which are written to a plurality of data in response to a plurality of write requests of the host, the method includes : The hard disk controller monitors a usage of the media cache; and In response to the usage of the media cache reaching a first threshold among a plurality of usage thresholds, the hard disk controller sends a first link packet to the host to request the host to A first delay time corresponding to the first threshold value is added between write requests. The method according to item 7 of the patent application scope, wherein the step of monitoring the usage of the media cache by the hard disk controller includes: The hard disk controller monitors the usage of the media cache based on a self-monitoring, analysis and reporting technology, and records the usage of the media cache in a vendor specific of the self-monitoring, analysis and reporting technology In the scratchpad. The method described in item 7 of the patent application scope further includes: In response to the usage of the media cache reaching a second threshold among the usage thresholds, the hard disk controller sends a second link packet to the host to request the host to A delay time is replaced with a second delay time corresponding to the second threshold value. The method as described in item 7 of the patent application scope, wherein the usage thresholds correspond to a plurality of delay times, the delay times include the first delay time, and the first threshold is positively related to the first delay. The method of claim 7 of the patent application, wherein the step of sending the first link packet to the host by the hard disk controller includes: The hard disk controller sends the first link packet to the host based on a device-initiated power management technology. The method described in item 7 of the patent application scope further includes: The hard disk controller moves the data to the main storage area in response to the first delay time between the write requests. A host computer connected to a stack of tiled magnetic recording hard disks, configured to: Issue multiple write requests to the shingled magnetic recording hard disk to write multiple data to a media cache of the shingled magnetic recording hard disk; Receiving a first link packet from the shingle-type magnetic recording hard disk, wherein the shingle-type magnetic recording hard disk reaches a first threshold among a plurality of usage thresholds in response to the usage of the media cache Value to request the host to add a first delay time between the write requests through the first link packet; Parse the first link packet and add the first delay time between the write requests accordingly.

Images