Storage device and management system equipped with ransomware attack detection function
By identifying and monitoring low-write-frequency pages for unusual activity, the storage device effectively detects ransomware attacks early, reducing damage through timely protective measures.
Patent Information
- Application Number
- JP2022164930
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2022-10-13
- Publication Date
- 2024-11-20
- Estimated Expiration
- 2042-10-13
AI Technical Summary
Conventional ransomware detection methods in storage devices are ineffective in detecting covert attacks that encrypt only part of the stored data, leading to delayed detection and extensive damage, as they rely on monitoring time-series changes in data entropy, compression rates, or deduplication rates, which are gradual and difficult to discern.
The storage device identifies low-write-frequency pages that have been updated a certain number of times within a period, marking them for potential ransomware attacks by analyzing access characteristics and thresholds, enabling early detection of tampering.
This approach allows for early detection of ransomware attacks, reducing damage by triggering protective measures such as data backups and snapshots, thereby minimizing the impact of such attacks.
Smart Images

Figure 0007587558000001 
Figure 0007587558000002 
Figure 0007587558000003
Abstract
Description
[Technical field]
[0001] The present invention generally relates to data processing performed by a storage device. [Background technology]
[0002] In recent years, cyber attacks targeting IT systems have been increasing. In particular, malware known as ransomware carries out cyber attacks that render IT systems unusable by encrypting or destroying data stored in the systems, and then demands a ransom in exchange for restoring the data. Damage caused by ransomware continues to occur both in Japan and overseas, and as a result, functions or solutions that increase the resilience of IT systems are attracting attention.
[0003] In particular, because storage devices store data for various IT systems, they are prone to cyber attacks that destroy data, such as ransomware. For this reason, technology is provided that detects data destruction by ransomware and protects data.
[0004] For example, Patent Document 1 discloses a technology in which a security appliance installed outside a storage device records operation events (the number of times a file is renamed, the number of times a file is updated, etc.) for monitored files stored in the storage device, and uses this monitoring information to detect ransomware attacks.
[0005] Furthermore, Patent Documents 2 and 3 disclose a storage device that monitors data entropy changes in write I / O of a host accessing the storage device and detects ransomware attacks based on the range of the changes.
[0006] Furthermore, Patent Document 4 discloses a technology in which decoy files for detecting ransomware are placed in the data storage area of a storage device, and changes due to data tampering in these decoy files are detected as a ransomware attack.
[0007] Furthermore, Patent Document 5 discloses a storage device that periodically takes snapshots of a data volume in the storage device and detects a ransomware attack when the amount of data difference between these taken snapshots or the difference in the compression rate of the volume meets a specified condition.
[0008] Furthermore, in view of the fact that ransomware attacks encrypt host I / O data and reduce the efficiency of data compression and data deduplication, Patent Document 6 discloses technology in which an appliance installed outside a storage device monitors host I / O and detects a ransomware attack when a change in the data compression rate or data deduplication rate over a certain period of time (sliding window) exceeds a threshold. [Prior art documents] [Patent documents]
[0009] [Patent Document 1] US10055582 [Patent Document 2] US10409986 [Patent Document 3] US10078459 [Patent Document 4] US10609066 [Patent Document 5] US2020 / 0042707 [Patent Document 6] US2020 / 0099699 Summary of the Invention [Problem to be solved by the invention]
[0010] Generally, the purpose of ransomware is not to destroy the IT system itself, but to obtain money by making the data stored in the IT system worthless. Therefore, an attack method is adopted that makes the stored data worthless without the user of the system realizing it. For example, some types of ransomware carry out a data destruction attack by excluding data necessary for the operation of the system from the target of the attack, and tampering with other data while continuing to operate the IT system itself normally. In addition, sophisticated processing implementations are often used to complete the attack in a short time, such as encrypting only a portion of the stored data.
[0011] For this reason, in actual ransomware attacks, it is often the case that a considerable amount of time has passed since the start of the attack before the damage actually becomes apparent or before the ransomware attack is addressed (i.e., the damage spreads to a wide area of IT systems).
[0012] Conventional ransomware attack detection technologies (disclosed in Patent Documents 1-6) mainly monitor time-series information such as changes in data entropy due to host I / O, fluctuations in the compression rate or data deduplication rate of stored data, and the amount of data change between periodically acquired data snapshots, and detect ransomware attacks from the amount of change. In other words, it is only after a ransomware attack has progressed to a certain extent that it becomes possible to detect the ransomware attack.
[0013] However, as mentioned above, ransomware attacks are carried out covertly by encrypting only a portion of stored data, and it is therefore believed that it is practically difficult to detect attacks using detection methods based on time-series changes in monitoring information, as in the conventional technology. In other words, because the amount of change in the time series is gradual, it is difficult to detect ransomware attacks even over time, and therefore the impact of damage caused by ransomware attacks may be large. [Means for solving the problem]
[0014] The storage device receives an access request from one of the hosts specifying one of one or more LDEVs (one or more logical volumes provided to one or more hosts), and in response to the access request, accesses a page assigned to the access destination area of the LDEV among a plurality of pages (a plurality of logical storage areas that can be assigned to one or more LDEVs).The storage device or its management system identifies mark target pages, which are low write frequency pages for which a certain number of writes have occurred within a certain period of time, based on management information including information indicating the write status characteristics of each of the plurality of pages, and checks for the possibility of a ransomware attack in which the number of mark target pages is equal to or exceeds a threshold value. Effect of the Invention
[0015] According to the exemplary embodiment of the present invention, it is possible to reduce the impact of damage caused by a ransomware attack. Problems, configurations, and effects other than those described above will become apparent from the following description of the embodiment. [Brief description of the drawings]
[0016] [Figure 1] 2 illustrates an example of the configuration of a storage device according to an embodiment. [Diagram 2] An example in which a ransomware attack detection function is provided in a storage device is shown. [Diagram 3] An example of a management system equipped with ransomware attack detection functionality is shown below. [Figure 4] 13 shows an example of the configuration of management information. [Diagram 5] 13 shows an example of the configuration of an LDEV management table. [Figure 6] 1 shows an example of the configuration of an address management table. [Figure 7] 13 shows an example of the configuration of a pool management table. [Figure 8] 1 shows an example of the configuration of a page management table. [Figure 9] 13 shows an example of the configuration of a pooled VOL management table. [Figure 10]1 shows an example of the configuration of a page access management table. [Figure 11] 1 shows an example of a ransomware attack detection process flow. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0017] An embodiment of the present invention will now be described with reference to the drawings.
[0018] Note that the embodiments described below do not limit the invention according to the claims, and not all of the combinations of elements described in the embodiments are necessarily essential to the solution of the invention.
[0019] In the following description, information that provides an output for an input may be described using expressions such as "xxx table" and "xxx list," but this type of information may be data of any structure, or may be a learning model such as a neural network that generates an output for an input. Therefore, "xxx table" and "xxx list" may be referred to as "xxx information." In the following description, the configuration of each table or list is an example, and one table or list may be divided into two or more tables or lists, or two or more tables or lists may all or partly be one table or list.
[0020] Although "numbers" are used as identification information for each piece of information, information other than numbers (for example, "first name", "name", "ID") can be used as identification information instead of or in addition to numbers. In the figures, numbers are represented as "#".
[0021] Furthermore, the embodiments of the present invention described below may be implemented as software running on a general-purpose computer, or may be implemented as dedicated hardware, or a combination of software and hardware.
[0022] In addition, in the following description, the processing may be described with the "program" as the subject, but since the program is executed by the processor unit to perform a defined process while appropriately using a memory unit and / or an interface unit, etc., the subject of the processing may be the processor unit (or a device such as a controller having the processor unit). The program may be installed in an apparatus such as a computer from a program source. The program source may be, for example, a program distribution server or a (e.g., non-transitory) recording medium readable by a computer. In the following description, two or more programs may be realized as one program, and one program may be realized as two or more programs.
[0023] In the following description, a "host system" refers to one or more host computers. At least one of the host computers may be a virtual computer running on a physical computer.
[0024] In the following description, a "management system" refers to one or more management computers. At least one management computer may be a virtual computer running on a physical computer. The management system may also include a management computer as an input / output console (for example, a management computer as a client).
[0025] In the following description, the "storage device" may be a storage device such as a so-called disk array device having a plurality of PDEVs, but may instead be a general-purpose computer (for example, a software-defined device such as SDS (Software-Defined Storage) or SDDC (Software-Defined Datacenter) that is realized by installing and executing software for achieving storage functions). Furthermore, the storage device may have a function as at least a part of a host in addition to the function as a storage device.
[0026] Furthermore, in the following explanation, "PDEV" refers to a physical storage device, and may typically be a persistent storage device. A PDEV may be, for example, an HDD (Hard Disk Drive) or an SSD (Solid State Drive). Different types of PDEVs may be mixed in a storage device.
[0027] In the following description, an "interface unit" may refer to one or more interface devices. The one or more interface devices may be at least one of the following: One or more I / O (Input / Output) interface devices. The I / O (Input / Output) interface devices are interface devices to at least one of the I / O devices and a remote display computer. The I / O interface device to the display computer may be a communications interface device. The at least one I / O device may be a user interface device, e.g., either an input device such as a keyboard and a pointing device, or an output device such as a display device. One or more communication interface devices. The one or more communication interface devices may be one or more homogeneous communication interface devices (e.g., one or more NICs (Network Interface Cards)) or two or more heterogeneous communication interface devices (e.g., a NIC and an HBA (Host Bus Adapter)).
[0028] In the following description, a "memory" refers to one or more memory devices, which are an example of one or more storage devices, and may typically be a primary storage device. At least one memory device in the memory may be a volatile memory device or a non-volatile memory device.
[0029] In the following description, a "persistent storage device" may be one or more persistent storage devices, which are an example of one or more persistent storage devices. A persistent storage device may typically be a non-volatile storage device (e.g., an auxiliary storage device), and more specifically, may be, for example, a hard disk drive (HDD), a solid state drive (SSD), a non-volatile memory express (NVME) drive, or a storage class memory (SCM).
[0030] Furthermore, in the following description, a "processor" may be one or more processor devices. The at least one processor device may typically be a microprocessor device such as a CPU (Central Processing Unit), but may also be other types of processor devices such as a GPU (Graphics Processing Unit). The at least one processor device may be a single-core or multi-core. The at least one processor device may be a processor core. The at least one processor device may also be a broader processor device such as a circuit that is a collection of gate arrays written in a hardware description language that performs part or all of the processing (e.g., an FPGA (Field-Programmable Gate Array), a CPLD (Complex Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit)).
[0031] In the following description, "RAID" is an abbreviation for Redundant Array of Inexpensive Disks. A RAID group may be an example of a persistent storage device, which is made up of multiple PDEVs (typically the same type of PDEVs) and stores data according to the RAID level associated with the RAID group. A RAID group may be called a parity group. A parity group may be, for example, a RAID group that stores parity.
[0032] Furthermore, in the following explanation, "volume (VOL)" refers to a storage area of a storage, and may be implemented by a physical storage device or a logical storage device.
[0033] Furthermore, a VOL may be a physical VOL (RVOL) or a virtual VOL (VVOL). An "RVOL" may be a VOL based on physical storage resources (e.g., one or more RAID groups) of a storage device having the RVOL. A "VVOL" may be any of an Externally Connected VOL (EVOL), a Capacity Expanded VOL (TPVOL), and a Snapshot VOL. An EVOL is based on the storage space (e.g., a VOL) of an external storage device and may be a VOL that complies with storage virtualization technology. A TPVOL is composed of multiple virtual areas (virtual storage areas) and may be a VOL that complies with capacity virtualization technology (typically Thin Provisioning).
[0034] A "storage pool (POOL)" is a logical storage area (for example, a collection of multiple pool VOLs), and may be prepared for each purpose. For example, there may be at least one type of storage pool, a TP pool, and a snapshot pool. A TP pool may be a storage area configured with multiple pages (a unit of data storage area divided by a certain size). When a page is not assigned to a virtual area (a virtual area of a TPVOL) to which an address specified in a write request received from a host system (hereinafter, a host) belongs, the storage controller assigns a page from the TP pool to that virtual area (a write destination virtual area) (even if a page has already been assigned to the write destination virtual area, a page may be newly assigned to the write destination virtual area). The storage controller may write the write target data associated with the write request to the assigned page. A snapshot pool may be a storage area in which data evacuated from an original VOL is stored. One pool may be used as both a TP pool and a snapshot pool. A "pool VOL" may be a VOL that is a component of a pool. A pool VOL may be an RVOL or an EVOL.
[0035] In the following explanation, a VOL recognized by a host (a VOL provided to a host by a storage device) is called an "LDEV." In the following explanation, an LDEV is a TPVOL (or RVOL), and a pool is a TP pool. However, the present invention can also be applied to storage devices that do not employ capacity expansion technology (Thin Provisioning).
[0036] In addition, in the following description, when describing elements of the same type without distinguishing between them, common reference symbols will be used, and when describing elements of the same type without distinguishing between them, reference symbols will be used.
[0037] FIG. 1 shows an example of the configuration of a storage device according to this embodiment.
[0038] One or more hosts 104 are connected to the storage device 100 via a network 108. A management system 105 is also connected to the storage device 100 via a management network 109. The network 108 is, for example, a Fibre Channel (FC) or an Internet Small Computer System Interface (iSCSI). The management network 109 is, for example, the Internet or a Wide Area Network (WAN). The network 108 and the management network 109 may be the same network.
[0039] Host 104 is an abbreviation for host system, and refers to one or more hosts. The host 104 has an HI / F (host interface device) 106, and transmits an access request (write request or read request) to the storage apparatus 100 and receives a response to the access request (for example, a write response including write completion, or a read response including the data to be read) via the HI / F 106. The HI / F 106 is, for example, an HBA (Host Bus Adapter) or a NIC (Network Interface Card).
[0040] The management system 105 manages the configuration and status of the storage apparatus 100. The management system 105 has an MI / F (management interface device) 107, and transmits commands to the storage apparatus 100 and receives responses to those commands via the MI / F. The MI / F 107 is, for example, a NIC.
[0041] Furthermore, the management system 105 may be software executed on a server or a PC that manages the storage device 100, or may be implemented as a function of a security appliance or software that manages the host 104 connected to the storage device 100. In addition to the MI / F 107, the management system 105 has a memory and a processor connected to the MI / F 107 and the memory.
[0042] The storage apparatus 100 has a plurality of drives 118, and a storage controller 103 connected to the plurality of drives 118. One or more RAID groups may be configured that include the plurality of drives 118. The drives 118 are an example of a PDEV.
[0043] The storage controller 103 includes a FI / F (front-end interface device) 110, a BI / F (back-end interface device) 117, a CM (cache memory) 111, an NVRAM (Non-Volatile RAM) 112, an MPPK 114A and an MPPK 114B, and a relay 113 that relays communications between these elements. The relay 113 is, for example, a bus or a switch.
[0044] The FI / F 110 is an I / F that communicates with the host 104 or a management server. The BI / F 117 is an I / F that communicates with the drive 118. The BI / F 117 may include an E / D circuit (a hardware circuit for encryption and decryption). Specifically, for example, the BI / F 117 may include a SAS (Serial Attached SCSI) controller, and the SAS controller may include the E / D circuit.
[0045] Data to be written to drive 118 or data read from drive 118 is temporarily stored in CM 111 (e.g., DRAM (Dynamic Random Access Memory)) by MPPK 114. Data in CM 111 (e.g., dirty data (data not written to drive 118)) is saved in NVRAM 112 by MPPK 114 which receives power from a battery (not shown) in the event of a power outage.
[0046] A cluster is configured by the MPPK 114A and the MPPK 114B. The MPPK 114A (MPPK 114B) has a DRAM 116A (DRAM 116B) and a CPU 115A (CPU 115B). The DRAM 116A (DRAM 116B) stores a control program 101A (control program 101B) executed by the CPU 115A (CPU 115B) and management information 102A (management information 102B) referenced or updated by the CPU 115A (CPU 115B). For example, I / O processing and address conversion processing of the storage device 100 are performed by the CPU 115A (CPU 115B) executing the control program 101A (control program 101B). At least one of the control program 101A (control program 101B) and the management information 102A (management information 102B) may be stored in a storage area (for example, the CM 111) shared by the multiple MPPKs 114A and 114B.
[0047] The FI / F 110 and the B-IF 117 may be examples of an interface device. The CM 111, the NVRAM 112, and the DRAM 116 may be examples of a memory. The drive 118 may be an example of a persistent storage device. The CPU 115 may be an example of a processor.
[0048] FIG. 2 shows an example in which a ransomware attack detection function is provided in a storage device 100.
[0049] The storage device 100 provides the LDEV 200 , and the host 104 transmits an access request specifying the LDEV 200 .
[0050] The actual state of the LDEV 200 is a VOL. The storage device 100 associates the storage area of the storage pool 201 with the storage area of the LDEV 200 provided to the host 104 in page 207 units.
[0051] Furthermore, the storage area constituting the storage pool 201 is provided by one or more pooled VOLs 202. The storage apparatus 100 manages the correspondence between the storage pool 201 and the pooled VOLs 202 that provide the storage area to the storage pool 201.
[0052] The storage apparatus 100 also has a page access monitoring program 203, a data protection control program 204, a ransomware attack detection program 205, and a threshold determination program 206. These programs 203 to 206 are stored in the DRAM 116A (DRAM 116B) and executed by the CPU 115A (CPU 115B).
[0053] The page access monitoring program 203 monitors I / O information for each page 207 constituting the storage pool 201. For each page 207, the I / O information includes information indicating at least one of the following: page allocation status (unassigned or allocated), I / O type (read or write), I / O pattern (sequential access, random access), number of read accesses, number of write accesses, amount of read data, and amount of write data.
[0054] By using the I / O information of each of these pages 207, the page access monitoring program 203 can classify the access characteristics of each page 207. For example, the page access monitoring program 203 can classify the access characteristics of a page 207 into any of the following access characteristics according to the access characteristics of the page 207: "HOT" means that the data is frequently accessed by the host 104 and frequently updated. "NORMAL" means that the host 104 accesses the data with a certain frequency and data updates occur. "COLD" means that data is accessed infrequently from the host 104 and updated. "ARCHIVE" is a state in which the host 104 rarely accesses the database and data is rarely updated.
[0055] Hereinafter, a page having access characteristics "KKK" will be referred to as a "KKK page." In this embodiment, each page 207 in the storage pool 201 is any one of a HOT page 207A, a NORMAL page 207B, a COLD page 207C, and an ARCHIVE page 207D. Furthermore, in this embodiment, the I / O information for each page 207 is information that includes access status information (information 1003 to 1008) in an entry of a page access management sub-table 1000 (see FIG. 10) described below. In other words, the access characteristics of the page 207 are determined based on the access status information of the page 207.
[0056] The storage device 100 performs classification according to the access characteristics of each page 207, and performs optimal page allocation corresponding to the performance characteristics of the pool VOLs 202 constituting the storage pool 201 according to the access characteristics of each page 207. This type of control is generally called "storage tiering". For example, storage tiering is as follows. That is, when drives 118 such as SSDs and HDDs with different access performances (and data storage costs) are mixed in one storage device 100, the access performances (and data storage costs) differ for each pool VOL 202. For this reason, by moving data in the page 207 allocated to the LDEV 200 to a page 207 based on the drive 118 with the access performance that is optimal for the characteristics of data access to the LDEV 200 by the host 104, the data storage efficiency in the storage device 100 can be improved.
[0057] Here, for example, consider a case where host 104A is infected with ransomware and data in LDEV 200A is encrypted (altered). Since the ransomware encrypts the data stored in LDEV 200A, data updates occur in pages 207 allocated to LDEV 200A. At this time, since the ransomware encrypts all pages 207 allocated to LDEV 200A, data updates are also performed, for example, in ARCHIVE page 207D (page 207 where data updates rarely occur).
[0058] Therefore, the ransomware attack detection program 205 detects I / O access that may be a ransomware attack (data tampering) by detecting when data updates occur to multiple ARCHIVE pages 207D assigned to LDEV 200A in the storage pool 201 during a certain period of time.
[0059] Furthermore, ransomware often infects multiple hosts 104 at the same time. For this reason, it is highly likely that data updates will occur at the same time for ARCHIVE page 207D allocated to LDEV200B and LDEV200C accessed from host 104B and host 104C in addition to host 104A. The ransomware attack detection program 205 detects that the frequency of data updates to ARCHIVE page 207D present in the storage pool 201 during a certain period of time has increased compared to normal times, as a possible sign of a ransomware attack.
[0060] When the ransomware attack detection program 205 detects the possibility of a ransomware attack (data tampering), as an example of a ransomware countermeasure, it notifies the storage administrator of abnormality detection information (information indicating abnormality detection) via the management network 109 and the management system 105. Furthermore, the data protection control program 204 performs a protection measure for stored data as an example of a ransomware countermeasure. This protection measure for stored data may be, for example, acquisition of a backup or snapshot of each LDEV 200, or creation of a restore point before the possible detection of ransomware from among the acquired backups or snapshots of each LDEV 200, thereby enabling easy data restoration. Furthermore, for example, some ransomware performs a delete operation of restore points or backups on the storage device 100, and the protection measure for stored data may be to prevent the backups or snapshots from being deleted by temporarily refusing these operations. Also, for example, protective measures for stored data may involve changing data update access (write I / O) from the host that occurs after detecting behavior that may be a ransomware attack to an I / O mode called Copy-on-Write (CoW) (an I / O mode in which the stored data before the update is temporarily evacuated to another area and then the stored data is updated), thereby preventing data tampering by ransomware.
[0061] In this manner, it is expected that damage caused by a ransomware attack can be reduced and recovery can be speeded up by detecting the possibility of a ransomware attack in the storage device 100 and taking measures to protect stored data. The threshold determination program 206 determines the threshold for the number of ARCHIVE pages 207D that satisfy a predetermined condition, as described below.
[0062] FIG. 3 shows an example in which the management system 105 is provided with a ransomware attack detection function.
[0063] A ransomware attack detection program 205, a data protection control program 204, and a threshold determination program 206 are implemented in a management system 105 installed outside the storage device 100, and executed by a processor of the management system 105.
[0064] Specifically, a page access monitoring program 203 in the storage device 100 periodically transmits I / O information 300 of each page 207 in the storage pool 201 to a ransomware attack detection program 205 in the management system 105 via the management network 109 .
[0065] The ransomware attack detection program 205 in the management system 105 detects I / O access that also causes data updates to the ARCHIVE pages 207D for each page 207 in the storage pool 201 based on the received I / O information 300. If the ransomware attack detection program 205 detects that data updates have occurred to multiple ARCHIVE pages 207D in the storage pool 201 during a certain period of time, it detects the possibility of a ransomware attack and notifies the storage administrator of anomaly detection information.
[0066] Furthermore, the data protection control program 204 in the management system 105 issues a protection instruction to the storage apparatus 100 via the management network 109. In response to this protection instruction, the storage apparatus 100 takes measures to protect the stored data.
[0067] In this way, by detecting the possibility of a ransomware attack on the storage device 100 and taking measures to protect stored data, it is expected that damage caused by a ransomware attack can be reduced and recovery can be speeded up.
[0068] This embodiment will be described in detail below. Note that in the following description, it is assumed that the example shown in FIG.
[0069] FIG. 4 shows an example of the configuration of management information in the storage device 100. As shown in FIG.
[0070] The management information 102 is composed of a plurality of management tables used by the control program 101 in the storage controller 103 in the storage apparatus 100. The management tables include, for example, an LDEV management table 401 that holds information about the LDEV 200, an address management table 404 that holds information about the correspondence between the addresses of areas in the LDEV 200 and the addresses of allocated pages 207, a pool management table 400 that holds information about the storage pool 201, a page management table 403 that holds information about managing pages 207 in the storage pool 201, a pool VOL management table 402 that holds information about pool VOL 202, and a page access management table 405 that holds information about the access status of the pages 207.
[0071] FIG. 5 shows an example of the configuration of the LDEV management table 401.
[0072] The LDEV management table 401 has an entry (record) for each LDEV 200. The information stored in each entry is an LDEV number 500, an LDEV capacity 501, a VOL type 502, and a pool number 503.
[0073] The LDEV number 500 indicates the identification number of the LDEV 200. The LDEV capacity 501 indicates the capacity of the LDEV 200. The VOL type 502 indicates the type of the LDEV 200 (for example, "RVOL" or "TPVOL"). The pool number 503 indicates the identification number of the storage pool 201 with which the LDEV 200 is associated. The LDEV 200 is assigned a page 207 from the storage pool 201 indicated by the pool number 503 corresponding to the LDEV 200.
[0074] FIG. 6 shows an example of the configuration of the address management table 404.
[0075] The address management table 404 is made up of a list 601 of LDEV numbers, and an address management sub-table 600 for each LDEV number. The address management sub-table 600 has an entry for each logical address range of the LDEV 200. The information stored in each entry is a top LBA 602, a pool number 604, and an allocated page number 603.
[0076] The first LBA 602 indicates the first LBA of the logical address range. The pool number 604 indicates the identification number of the storage pool 201 having the page 207 allocated to the logical address range. The allocated page number 603 indicates the identification number of the allocated page 207.
[0077] FIG. 7 shows an example of the configuration of the pool management table 400.
[0078] The pool management table 400 has an entry for each storage pool 201. The information stored in each entry is a pool number 703, a pool capacity 700, a pool allocated capacity 701, and a pool used capacity 702.
[0079] The pool number 703 indicates the identification number of the storage pool 201. The pool capacity 700 indicates the defined capacity of the storage pool 201 (specifically, the sum of one or more VOL capacities corresponding to one or more pooled VOLs 202 constituting the storage pool 201). The pool allocated capacity 701 indicates the actual capacity allocated to one or more LDEVs 200 associated with the storage pool 201 (specifically, the total capacity of all pages 207 allocated to one or more LDEVs 200). The pool used capacity 702 indicates the total amount of data stored in the storage pool 201.
[0080] FIG. 8 shows an example of the configuration of the page management table 403.
[0081] The page management table 403 is made up of a list 801 of pool numbers and a page management sub-table 800 for each pool number. The page management sub-table 800 has an entry for each page 207 in the storage pool 201. The information stored in each entry is a page number 802, an allocation status 804, an access characteristic 806, a pool VOL number 805, and a pool VOL first LBA 803.
[0082] The page number 802 indicates the identification number of the page 207. The allocation status 804 indicates the allocation status of the page 207, for example, "allocated" if it has been allocated, and "not allocated" if it has not been allocated. The access characteristics 806 indicate the access characteristics of the page 207 (specifically, the above-mentioned "HOT", "COLD", "ARCHIVE", or "NORMAL"). The access characteristics are determined by the page access monitoring program 203 based on the I / O status from the host 104 for a certain period of time. The pool VOL number 805 indicates the identification number of the pool VOL 202 which has the page 207. The pool VOL first LBA 803 indicates the first LBA of the page 207.
[0083] FIG. 9 shows an example of the configuration of the pooled volume management table 402.
[0084] The pooled volume management table 402 has a pool number list 910, and a pooled volume sub-table 900 for each pool number. The pooled volume sub-table 900 has an entry for each pooled volume 202. The information stored in each entry is a pooled volume number 905, a PDEV type 901, and a pooled volume capacity 902.
[0085] The pooled VOL number 805 indicates the identification number of the pooled VOL 202. The PDEV type 901 indicates the type of PDEV (drive 118 in this embodiment) that is the basis of the pooled VOL 202. The pooled VOL capacity 902 indicates the capacity of the pooled VOL 202.
[0086] FIG. 10 shows an example of the configuration of the page access management table 405.
[0087] The page access management table 405 is a table that holds information regarding the access status of each page 207 in the storage pool 201. The page access monitoring program 203 monitors the access status of each page 207 and updates the table 405.
[0088] The page access management table 405 has a list of pool numbers 1010, and a page access management sub-table 1000 for each pool number. The page access management sub-table 1000 has an entry for each page 207 that constitutes the storage pool 201. The information stored in each entry is a page number 1002, a timestamp 1001, a read I / O count 1003, a write I / O count 1004, a sequential access ratio 1005, a random access ratio 1006, a write data amount 1007, and a read data amount 1008.
[0089] The page number 802 indicates the identification number of the page 207. The timestamp 1001 indicates a timestamp indicating the timing at which the page 207 was accessed. The read I / O count 1003 indicates the number of read I / Os to the page 207. The write I / O count 1004 indicates the number of write I / Os to the page 207. The sequential access ratio 1005 indicates the number of sequential accesses among the accesses (I / Os) to the page 207. The random access ratio 1006 indicates the number of random accesses among the accesses (I / Os) to the page 207. The write data amount 1007 indicates the cumulative amount of data written to the page 207. The read data amount 1008 indicates the cumulative amount of data read from the page 207.
[0090] Figure 11 shows an example of a ransomware attack detection process flow.
[0091] The ransomware attack detection process may be performed periodically, or in response to a predetermined event such as an update of the page access management table 405. The ransomware attack detection process is asynchronous with the I / O process in response to an access request specifying the LDEV 200, but may also be synchronous with the I / O process.
[0092] In step 1101, the ransomware attack detection program 205 refers to the pool management table 400 and selects the pool number of the storage pool 201 to be detected from the pool number list 801. The "storage pool 201 to be detected" may be the storage pool 201 corresponding to the updated page access management sub-table 1000, or may be the storage pool 201 that has become the order of the ransomware attack detection process in a predetermined order such as round robin.
[0093] In step 1102, the ransomware attack detection program 205 refers to the page management sub-table 800 corresponding to the storage pool 201 to be detected, and extracts all page numbers 802 whose access characteristics 806 are "ARCHIVE". The ransomware attack detection program 205 selects one page number 802 from the extracted page numbers 802.
[0094] In step 1103, the ransomware attack detection program 205 refers to the page access management sub-table 1000 corresponding to the storage pool 201 to be detected, and obtains an entry containing a page number 1002 that matches the page number 802 selected in step 1102 (or step 1107).
[0095] In step 1104, the ransomware attack detection program 205 refers to the timestamp 1001 from the entry acquired in step 1103, and checks whether the time represented by the timestamp 1001 (for example, the last update time of the page 207 corresponding to the entry) is less than a certain period from the current time (i.e., less than a threshold). If the check result is false, the process proceeds to step 1107. On the other hand, if the check result is true, the process proceeds to step 1105.
[0096] In step 1105, the ransomware attack detection program 205 refers to the write I / O count 1004 from the entry acquired in step 1103, and checks whether the write count indicated by the write I / O count 1004 is equal to or greater than a threshold. If the check result is true, the process proceeds to step 1106. On the other hand, if the check result is false, the process proceeds to step 1107.
[0097] In step 1106, the ransomware attack detection program 205 marks the page as an evaluation target for the detection process.
[0098] In step 1107, the ransomware attack detection program 205 selects the next page number 802 from the page numbers 802 extracted in step 1102.
[0099] In step 1108, the ransomware attack detection program 205 checks whether the next page number 802 was selected in step 1107. If the check result is true, the process returns to step 1103. If the check result is false, the process proceeds to step 1109.
[0100] In step 1109, the ransomware attack detection program 205 checks whether the number of pages marked as targets for evaluation in the detection process in step 1106 is equal to or greater than a threshold. If the check result is true, then a ransomware attack may be occurring, since data updates have occurred intensively for a certain number or more of ARCHIVE pages 207D over a certain period of time (data updates have occurred for many ARCHIVE pages 207D that are not normally accessed). In this case, processing proceeds to step 1110. On the other hand, if the check result is false, processing proceeds to step 1111.
[0101] In step 1110, the ransomware attack detection program 205 notifies the management system 105 of the anomaly detection information. The notification of the anomaly detection information is an example of a protective measure for stored data.
[0102] In step 1111, the ransomware attack detection program 205 checks whether there is an unprocessed storage pool 201 to be detected. If the check result is true, the process returns to step 1101. If the check result is false, the process ends.
[0103] Although one embodiment has been described above, this is merely an example for explaining the present invention, and the scope of the present invention is not limited to this embodiment. The present invention can be embodied in other forms.
[0104] The above description can be summarized, for example, as follows. The following summary may include supplementary explanations and explanations of modifications of the above description.
[0105] The storage device 100 receives an access request specifying one of one or more LDEVs 200 (one or more logical volumes provided to one or more hosts 104) from one of the hosts 104, and in response to the access request, accesses one of multiple pages 207 (multiple logical storage areas that can be allocated to one or more LDEVs 200) that is allocated to the access destination area of the LDEV. For example, the following may be performed. The control program 101 refers to the address management sub-table 600 corresponding to the LDEV 200 specified in the access request, and identifies whether or not a page 207 has been allocated to the access destination area of the LDEV 200 . If the access request is a write request and a page 207 has not been assigned, the control program 101 identifies the pool number corresponding to the specified LDEV 200 from the LDEV management table 401, identifies an unallocated page (free page) 207 from the page management sub-table 800 corresponding to the identified pool number, and assigns the identified page 207 to the access destination area in the LDEV 200. The control program 101 accesses the allocated page 207. Furthermore, the control program 101 updates, as necessary, an entry in the pool management table 400 that corresponds to the pool that has the accessed page 207. The page access monitoring program 203 updates the access status (at least a part of the information 1001 to 1008) of the accessed page 207.
[0106] The ransomware attack detection program 205 identifies mark target pages that are low write frequency pages for which a certain number of writes have occurred within a certain period of time, based on the management information 102 including information indicating the write status characteristics for each of the multiple pages 207. A "low write frequency page" is a page among the multiple pages 207 that corresponds to a low write frequency defined as having a relatively low write frequency based on the write status characteristics. The ransomware attack detection program 205 checks for the possibility of a ransomware attack in which the number of mark target pages is equal to or greater than a threshold value.
[0107] The occurrence of data updates in two or more pages where data updates do not normally occur indicates the possibility of data tampering or erroneous data manipulation by ransomware, that is, the possibility of a ransomware attack, and this possibility can be detected. As a result, protective measures can be taken to prevent damage from becoming too great, thereby reducing the impact of damage from a ransomware attack. In addition, for example, ransomware is considered to often infect multiple hosts 104 at the same time, and therefore it is considered highly likely that data updates will occur at the same time for multiple low write frequency pages assigned to multiple LDEVs 200 accessed by multiple hosts, but the ransomware attack detection program 205 can detect the possibility of a ransomware attack when the data update frequency for low write frequency pages present in the storage pool 201 during a certain period of time becomes higher than normal.
[0108] The management information 102 may include the time of write each time a write occurs for each page 207, or may include information indicating the number of writes since a certain time, and it is possible to identify low write frequency pages for which a certain number of writes or more have occurred within a certain period of time from such management information 102. Furthermore, the "write status characteristics" may be represented by at least one of information indicating the write frequency (for example, the number of write I / Os since a certain time 1004) and the access characteristics 806.
[0109] Furthermore, an example of a low write frequency page is the ARCHIVE page 207D, but instead of the ARCHIVE page 207D, a page with other access characteristics, such as the COLD page 207C, may also be an example of a low write frequency page.
[0110] If the result of the above check is false, at least one of the following measures may be taken to mitigate the effects of ransomware attacks. (a) When the target LDEV 200, which is an LDEV 200 to which a low write frequency page to which writing has occurred within a certain period of time is assigned, is the data write destination, and a write occurs to update data in a page 207 assigned to the target LDEV 200, the data before the update is evacuated (for example, Copy-On-Write). (b) When an instruction to delete a snapshot of the target LDEV 200 is received, the instruction is rejected. (c) When the target LDEV 200 has a remote copy destination VOL (logical volume) and data in page 207 assigned to the target LDEV is remote copied to the remote copy destination VOL, the data before the update in the remote copy destination VOL is evacuated (for example, the remote copy destination storage device of storage apparatus 100 is caused to execute the evacuation), or a snapshot of the remote copy destination VOL is obtained in, for example, the remote copy destination storage device and then the data is copied to the remote copy destination VOL. (d) Notifying the storage administrator of abnormality detection information, which is information indicating the detection of an abnormality. (e) In addition to the target LDEV 200, at least one of (a) through (d) is also performed for each of one or more LDEVs 200 other than the target LDEV 200 that belong to the same group (typically the same host group) as the target LDEV 200.
[0111] At least (a) may be performed. Furthermore, (a), (b) and (c) may be performed by the data protection control program 204. (d) may be performed by the ransomware attack detection program 205. With regard to (e), one or more LDEVs 200 other than the target LDEV 200 that belong to the same group (typically the same host group) as the target LDEV 200 may be identified from management information. For example, the data protection control program 204 may autonomously protect stored data, thereby creating at least one of backup data, restore points and snapshots of the target LDEV 200.
[0112] The management information 102 may include a page access management table 405 (an example of page access management information) including information indicating an access status for each of the multiple pages 207, and a page management table 403 (an example of page management information) including information indicating an access characteristic for each of the multiple pages 207. For each of the multiple pages 207, the access status of the page 207 may include the number of writes to the page (for example, the number of writes in a certain period). For each of the multiple pages 207, the access characteristic of the page may be a characteristic determined based on the write frequency of the page. The page access monitoring program 203 may update the access status corresponding to the page 207 represented by the page access management table 405 every time the page 207 is accessed. The ransomware attack detection program 205 may update the access characteristic 806 of the page represented by the page management table 403 based on the write frequency in the access status of the page, for each of the multiple pages 207, at a frequency lower than the update of the access status of the page 207. In this way, the access characteristics 806 can be made to follow updates to the access situation.
[0113] The multiple pages 207 are pages 207 based on multiple storage device groups with different access performance. Each of the multiple pages 207 may be based on one of the multiple storage device groups. Each storage device group may be a RAID group and may be one or more storage devices. The control program 101 may perform storage tiering. When the access status of the page 207 is not suitable for the access performance of the storage device group that is the basis of the page (for example, the pool VOL 202 based on the storage device group), the storage tiering may include moving data in the page 207 to a page based on a storage device group that has access performance suitable for the access status of the page 207 (for example, a page in the pool VOL 202 based on the storage device group). In storage tiering, for example, the control program 101 refers to the page access management table 405 and the pool VOL management table 402 to identify a page in the pool VOL 202 that has access performance suitable for the access status of the page 207. For this reason, at least a part of the information used in storage tiering can be effectively used to detect the possibility of a ransomware attack.
[0114] The threshold determination program 206 may determine the threshold for the number of pages to be marked based on, for example, statistics of the number of writes of low write frequency pages during a certain period and the number of low write frequency pages. An example of the statistics may be a distribution based on the relationship between the number of low write frequency pages and the number of writes during a certain period (for example, a distribution in which the horizontal axis is the number of writes during a certain period and the vertical axis is a probability density based on the number of low write frequency pages). The threshold for pages to be marked may be determined based on a condition that the number of low write frequency pages that have been written (updated) during a certain period falls within X% of the total. In addition to the above-mentioned method, this statistical processing may use, for example, a method that uses a machine learning algorithm to learn and use the distribution during normal operation. This optimizes the threshold, and thus improves the accuracy of detecting the possibility of a ransomware attack. For example, the threshold determination program 206 may determine the threshold for the number of marked pages by inputting the number of writes of low write frequency pages during a certain period and the number of low write frequency pages into a machine learning model (specifically, a machine learning model that takes the number of writes of low write frequency pages and the number of low write frequency pages as input and the threshold for marked pages as output). The machine learning model may be a neural network or another machine learning model.
[0115] Each of the ransomware attack detection program 205, the data protection control program 204, and the threshold determination program 206 may be in the storage apparatus 100 or in the management system 105. When the management system 105 has at least the ransomware attack detection program 205, at least a part of the management information 102 (for example, information contained in the page management table 403 and the page access management table 405) may be stored in the memory of the management system 105, and based on this information, the ransomware attack detection program 205 can check whether there is a possibility of a ransomware attack, and the threshold determination program 206 can determine the threshold for the number of pages to be marked.
[0116] Furthermore, the page 207 may be an area in an LDEV 200 (for example, an RVOL) instead of an area in the storage pool 201. [Explanation of symbols]
[0117] 100 Storage Device
Claims
1. a memory that stores management information including information that indicates a write status characteristic for each of a plurality of pages that are a plurality of logical storage areas that can be allocated to one or a plurality of LDEVs; a processor that receives an access request specifying one of the one or more LDEVs from one of the one or more hosts, and in response to the access request, accesses a page, of the plurality of pages, that is allocated to an area to be accessed by the LDEV; Equipped with The one or more LDEVs are one or more logical volumes provided to one or more hosts, Among the plurality of pages, a page corresponding to a low write frequency defined as having a relatively low write frequency based on a write status characteristic is a low write frequency page, The processor, Based on the management information, a mark target page is identified, which is a low write frequency page in which a certain number of writes have occurred within a certain period of time; Checking whether there is a possibility of a ransomware attack, in which the number of marked pages is equal to or greater than a threshold value; Storage device.
2. If the check is false, the processor performs at least one of the following actions to combat the ransomware: (a) when a target LDEV, which is an LDEV to which a low write frequency page to which a write has occurred during the certain period is assigned, is a data write destination and a write occurs to update data in a page assigned to the target LDEV, the data before the update is evacuated; (b) when an instruction to delete a snapshot of the target LDEV is received, the instruction is rejected; (c) when the target LDEV has a remote copy destination logical volume and data in a page allocated to the target LDEV is remote copied to the remote copy destination logical volume, either saving data before updating in the remote copy destination logical volume, or obtaining a snapshot of the remote copy destination logical volume and then copying the data to the remote copy destination logical volume; (d) notifying a storage administrator of abnormality detection information, which is information indicating the abnormality detection; (e) performing at least one of (a) to (d) on each of one or more LDEVs other than the target LDEV that belong to the same group as the target LDEV, in addition to the target LDEV; The storage device according to claim 1.
3. If the check is false, the processor performs at least (a). The storage device according to claim 2.
4. The management information is page access management information including information indicating an access status for each of the plurality of pages; page management information including information representing access characteristics for each of the plurality of pages; Including, For each of the plurality of pages, the access status of the page includes a number of writes to the page; For each of the plurality of pages, an access characteristic of the page is determined based on a write frequency of the page, the processor is configured to update an access status corresponding to a page, which is represented by the page access management information, every time the page is accessed; the processor updates, for each of the plurality of pages, the access characteristics of the page, which are represented by the page management information, based on a write frequency in the access status of the page, at a frequency lower than the update frequency of the access status of the page. The storage device according to claim 1.
5. the plurality of pages are pages based on a plurality of storage device groups having different access performances, Each storage device group is one or more storage devices, each of the plurality of pages is based on one of the plurality of storage device groups; The processor performs storage tiering; the storage tiering includes, when an access status of a page is not suitable for the access performance of a storage device group that is a base of the page, moving data in the page to a page based on a storage device group that has access performance suitable for the access status of the page; The storage device according to claim 4.
6. The processor determines the threshold value based on statistics of the number of writes of low write frequency pages during the certain period and the number of low write frequency pages. The storage device according to claim 1.
7. The processor determines the threshold by inputting the number of writes of low write frequency pages during the certain period and the number of low write frequency pages into a machine learning model; The machine learning model is a model in which the number of writes of low write frequency pages and the number of low write frequency pages are input, and the number of marked pages is a threshold value as an output. The storage device according to claim 1.
8. A storage device management system, a memory that stores management information including information that indicates a write status characteristic for each of a plurality of pages that are a plurality of logical storage areas that can be allocated to one or a plurality of LDEVs; a processor coupled to the memory; Equipped with The one or more LDEVs are one or more logical volumes provided to one or more hosts, the storage device receives an access request specifying one of the one or more LDEVs from one of the one or more hosts, and in response to the access request, accesses a page of the plurality of pages that is allocated to an area to be accessed by the LDEV, Among the plurality of pages, a page corresponding to a low write frequency defined based on a write status characteristic is a low write frequency page, The processor, Based on the management information, a mark target page is identified, which is a low write frequency page in which a certain number of writes have occurred within a certain period of time; Checking whether there is a possibility of a ransomware attack, in which the number of marked pages is equal to or greater than a threshold value; Management system.
9. 1. A method for managing a storage device, comprising: receiving an access request from one or more hosts, the access request designating one of one or more LDEVs which are one or more logical volumes provided to the one or more hosts; and, in response to the access request, accessing a page allocated to an area to be accessed by the LDEV, among a plurality of pages which are a plurality of logical storage areas which can be allocated to the one or more LDEVs, the method comprising: the computer identifies, based on management information including information representing a write status characteristic for each of the plurality of pages, a mark target page that is a low write frequency page for which a certain number of writes have occurred within a certain period of time; The low write frequency page is a page that corresponds to a low write frequency defined based on a write status characteristic, The computer checks for the possibility of a ransomware attack in which the number of marked pages is equal to or greater than a threshold value; Management method.
Citation Information
Patent Citations
Method for monitoring file access, program, and system
JP2019075131A
Data processing method for dealing with ransomware, program for executing the method, and computer-readable recording medium on which the program is recorded
JP2021528773A
Automated detection and remediation of ransomware attacks involving a storage device of a computer network
US10055582B1
Ransomware detection using I / O patterns
US10078459B1
Ransomware detection in a continuous data protection environment
US10409986B1
