BACKGROUND

[0001] Datacenters are widely used to store data for various entities and organizations. For example, large enterprises may store data at one or more datacenters that may be centrally located. The data stored at such datacenters may be vital to the operations of the enterprise and may include, for example, employee data, customer data, etc.

BRIEF DESCRIPTION OF THE DRAWINGS

[0002] For a more complete understanding of various examples, reference is now made to the following descriptions taken in connection with the accompanying drawings in which:

[0003] Figure 1 illustrates a system in accordance with an example;

[0004] Figure 2 illustrates the example system of Figure 1 forming two three-datacenter (3DC) arrangements; and

[0005] Figure 3 is a flow chart illustrating an example method. DETAILED DESCRIPTION

[0006] Various examples described herein provide for successful disaster recovery for datacenters with data currency, while also providing prevention of propagation of data corruption. In this regard, the present disclosure provides a solution which may be implemented as a four-data center arrangement which simultaneously forms a first three-datacenter (3DC) arrangement for replication of the primary data volume and a second 3DC arrangement for replication of a recovery data volume. The recovery data volume may lag in time, thereby preventing propagation of data corruption which may occur in the primary data volume.

[0007] Referring now to Figure 1 , a system in accordance with a first example is illustrated. The example system 100 of Figure 1 includes four data centers 110, 120, 130, 140. The first data center (Data Center A) 110 and the second data center (Data Center B) 120 may be located close to one another, while the third data center (Data Center C) 130 and fourth data center (Data Center D) 140 may be located close to one another, but away from the first data center 110 and the second data center 120. In this regard, in various examples, the first data center 110 and the second data center 120 may be located in a first region (e.g., same neighborhood, city, state, etc.), while the third data center (Data Center C) 130 and the fourth data center (Data Center D) 140 are located in a second region. Thus, if a disaster (e.g., natural, political, economic, etc.) were to strike the first region, operations of the enterprise may continue using the data centers located in the second region.

[0008] In the example illustrated in Figure 1, the first data center 110 includes a primary data volume 112 for storage of data. In various examples, the primary data volume 112 may include any of a variety of non-transitory storage media, including hard drives, flash drives, etc. Further, the data stored on the primary data volume 112 may include any type of data, including databases, program data, software programs, etc. The primary data volume 112 of the first data center 110 may be the primary storage of data for the enterprise. In this regard, the primary data volume 112 of the first data center 110 may be the storage source accessed for all data read and write requests.

[0009] In some examples, as illustrated in Figure 1, the first data center 110 may be provided with a second data volume, shown in Figure 1 as a recovery data volume 118. Those skilled in the art will appreciate that the primary data volume 112 and the recovery data volume 118 may be either separate storage media or separate portions (e.g., virtual) of a single storage medium.

[0010] The first data center 110 may also include various other components, such as a cache 114 and a server 116, which may include a processor and a memory. The cache 114 and the server 116 may facilitate in the storage, writing and access of the data on the primary data volume 112.

[0011] The second data center 120, the third data center 130, and the fourth data center 140 may each be provided with similar components as the first data center 110. For example, the second data center 120 includes a primary data volume 122, a cache 124 and a server 126. The example third data center 130 includes a primary data volume 132, a cache 134, a server 136 and a recovery data volume 138. Further, the example fourth data center 140 includes a recovery data volume 148, a cache 144 and a server 146.

[0012] As noted above, in various examples, the first data center 110 and the second data center 120 are located close to each other. As illustrated in Figure 1 , the primary data volume 112 of the first data center 110 and the primary data volume 122 of the second data center 120 replicate each other, as indicated by the arrow labeled "A". In various examples, when the first data center 110 and the second data center 120 are in relative close proximity to each other, the replication indicated as A is performed synchronously. In this regard, the server 116 may write data to the primary data volume 112 through the cache 114. The primary data volume 112 of the first data center 110 may write to the primary data volume 122 of the second data center 120. Only after the primary data volume 112 of the first data center receives an acknowledgement from the primary data volume 122 of the second data center 120, the write may be acknowledged to the server 116 of the first data center 110. Thus, synchronous replication may assure that all data written to the primary data volume 112 of the first data center 110 is replicated to the primary data volume 122 of the second data center 120. Synchronous replication is effective when the replicated data volumes are located relatively close to one another (e.g., less than 100 miles).

[0013] Referring again to Figure 1, the primary data volume 112 from the first data center 110 is replicated to the primary data volume 132 at the third data center 130, as indicated by the arrow labeled "B". In various examples, when the third data center 130 is located relatively far from the first data center 110 and the second data center 120, the replication indicated as B is performed asynchronously. In this regard, the server 116 may write data to the primary data volume 112 through the cache 114, and the write is immediately acknowledged to the server 116 of the first data center 110. The cached write data may then be pushed to the primary data volume 132 of the third data center 130. In other examples, the cached write data is indicated in a journal which may be stored within the first data center 110. The third data center 130 may periodically poll the first data center to read journal information and, if needed, retrieve data for writing to the primary data volume 132 of the third data center 130. Thus, since the

acknowledgment of the write is sent to the server 112 before any transfer of data to the third data center 130, asynchronous replication may not assure that all data written to the primary data volume 112 of the first data center 110 is replicated to the primary data volume 132 of the third data center 130. Asynchronous replication can be effectively implemented for data centers that are located far apart from one another (e.g., more than 100 miles).

[0014] While Figure 1 illustrates that the primary data volume 132 of the third data center 130 is asynchronously replicated from the first data center 110 (line B), in other examples, the primary data volume 132 of the third data center 130 may be replicated from the second data center 120. [0015] Referring again to Figure 1, at the third data center 130, the primary data volume 132 is used to generate, update or synchronize a recovery data volume 138, as indicated by the arrow labeled "C". In various examples, the recovery data volume 138 may be generated during an initial copy step. Thereafter, the recovery data volume 138 may be updated or synchronized with the primary data volume 132 at regular intervals, for example. While Figure 1 illustrates the recovery data volume 138 as being a separate storage device from the primary data volume 132, in various examples, the recovery data volume 138 may be provided on the same storage device in, for example, a virtual separated portion of the storage device.

[0016] In various examples, the generation of the recovery data volume 138 may be performed periodically based on, for example, a predetermined schedule. The frequency at which the recovery data volume 138 is generated may be determined based on the needs of the particular implementation. In one example, the recovery data volume 138 is generated every six hours.

[0017] Referring again to Figure 1, the recovery data volume 138 from the third data center 130 is replicated to the recovery data volume 148 at the fourth data center 140, as indicated by the arrow labeled "D". In various examples, the third data center 130 and the fourth data center 140 are located relatively close to one another. Thus, the replication of the recovery data volume 138 to the recovery data volume 148 may be effectively performed synchronously.

[0018] In some examples, the recovery data volume 138 of the third data volume 130 may be replicated to the recovery data volume 118 at the first data center 110, as indicated by the arrow labeled "E". In various examples, the third data center 130 and the first data center 110 are located relatively far from one another. Thus, the replication of the recovery data volume 138 to the recovery data volume 118 may be effectively performed asynchronously. While Figure 1 illustrates the replication of the recovery volume 118 of the first data center 110 from the third data center 130, in other examples, the recovery volume 118 may be replicated from the recovery volume 148 of the fourth data center 140.

[0019] The example of Figure 1 illustrates each data center 110, 120, 130, 140 provided with a server and a cache. Those skilled in the art will appreciate that, in some examples, the server may not be required in certain data centers. For example, the second data center 120 of Figure 1 may not require the server 126 under normal operation. The second data center 120 may only require the server 126 during a recovery mode in, for example, the event of a disaster during which the second data center 120 may be required to serve as the primary data center. Similarly, the third data center 130 and the fourth data center 140 may also not require a server in normal operation.

[0020] Referring now to Figure 2, the example system 100 of Figure 1 is illustrated as forming two three-datacenter (3DC) arrangements. In this regard, the first 3DC arrangement 210 is illustrated by the solid lined triangle. The first 3DC arrangement 210 includes the first data center 110, more particularly, the primary data volume 112 of the first data center 110. The first 3DC arrangement 210 is further formed by the second data center 120 and the third data center (more particularly, the primary data volume 132 of the third data center 130). As described above with reference to Figure 1, the second data center 120 includes the primary data volume 122 having a synchronous replication of the primary data volume 112 from the first data center 110. Further the third data center 130 includes an asynchronous replication of the primary data volume 112 from the first data center 110.

[0021] The example system 100 further includes a second 3DC arrangement 220, as illustrated by the dashed lined triangle. The second 3DC arrangement 220 includes the third data center 130, more particularly, the recovery data volume 138 of the third data center 130. The second 3DC arrangement 220 is further formed by the fourth data center 140 and the first data center (more particularly, the recovery data volume 118 of the first data center 110). As described above with reference to Figure 1, the fourth data center 140 includes the recovery data volume 148 having a synchronous replication of the recovery data volume 138 from the third data center 130. Further the first data center 110 includes an asynchronous replication of the recovery data volume 138 from the third data center 130.

[0022] Thus, the first data center 110 may include the primary data volume 112 having current data, as well as a recovery data volume having time-lagging data. Further the replicated recovery data volume is isolated from the primary data volumes. Therefore, protection against a local disaster, which may affect the entire region in which both the first data center 110 and second data center 120 are located, as well as protection against propagation of data corruption, is provided.

[0023] Referring now to Figure 3, a flow chart illustrating an example method is provided. In accordance with the example method 300, primary data volumes at the first data center 110 and the second data center 120 are synchronously replicating (block 310), as illustrated by the line labeled "A" in Figures 1 and 2 above. As described above, since the first and second data centers are located relatively in close proximity, synchronous replication is an effective mode.

[0024] The primary data volume at the first data center 110 is asynchronously replicated to the third data center (block 312), as illustrated by the line labeled "B" in Figures 1 and 2 above. Further, in examples where the third data center is located relatively far from the first data center, asynchronous replication is the most effective mode.

[0025] At block 314, a determination is made as to whether the time for periodic update or synchronization of the recovery volume has arrived. As noted above, the recovery volume may be initially generated as a copy of the primary volume and may be updated or synchronized with the primary volume on a periodic basis. If the determination is made that the time for updating or synchronization of the recovery volume has not yet arrived, the process returns to block 310 and continues synchronous replication of the first and second data centers and the asynchronous replication of the third data center. When the time for update or synchronization of the recovery volume has arrived, the process proceeds to block 316, and a recovery data volume is updated or synchronized with the primary volume at the third data center 130, as illustrated by the line labeled "C" in Figures 1 and 2. As noted above, the frequency at which the recovery data volume is updated or synchronized may be set for the particular implementation.

[0026] The recovery data volume at the third data center may then be synchronously replicated to the fourth data center (block 318), as illustrated by the line labeled "D" in Figures 1 and 2. Since the third and fourth data centers are located in close proximity to each other, synchronous replication can be effectively achieved.

[0027] The recovery data volume at the third data center 130 is asynchronously replicated to the first data center 110 (block 320), as illustrated by the line labeled "E" in Figures 1 and 2 above. Further, in examples where the third data center is located relatively far from the first data center, asynchronous replication is the most effective mode. The process 300 then returns to block 310.

[0028] Referring again to block 314, as noted above, when the determination is made that the time for update or synchronization of the recovery volume has arrived, the process proceeds to block 316 and updates or synchronizes the recovery data volume at the third data center. Those skilled in the art will understand that the updating or synchronization of the recovery data volume may be performed at the same time as the replication in blocks 310 and 312. Thus, the replication in blocks 310 and 312, which may be substantially continuous, need not be suspended during the updating or synchronization of the recovery data volume.

[0029] Thus, in accordance with various examples described herein, four data centers may be used to form two separate 3DC arrangements. One of the 3DC arrangements provides replication of the primary data of a data center at a nearby data center (synchronously) and a distant data center (asynchronously), while the other 3DC arrangement provides similar replication for a recovery data volume which lags in time. Thus, data protection is provided against regional disasters, as well as propagation of data corruption.

[0030] Various examples described herein are described in the general context of method steps or processes, which may be implemented in one example by a software program product or component, embodied in a machine-readable medium, including executable instructions, such as program code, executed by entities in networked environments. Generally, program modules may include routines, programs, objects, components, data structures, etc. which perform particular tasks or implement particular abstract data types. Executable instructions, associated data structures, and program modules represent examples of program code for executing steps of the methods disclosed herein. The particular sequence of such executable instructions or associated data structures represents examples of corresponding acts for implementing the functions described in such steps or processes.

[0031] Software implementations of various examples can be accomplished with standard programming techniques with rule-based logic and other logic to accomplish various database searching steps or processes, correlation steps or processes, comparison steps or processes and decision steps or processes.

[0032] The foregoing description of various examples has been presented for purposes of illustration and description. The foregoing description is not intended to be exhaustive or limiting to the examples disclosed, and modifications and variations are possible in light of the above teachings or may be acquired from practice of various examples. The examples discussed herein were chosen and described in order to explain the principles and the nature of various examples of the present disclosure and its practical application to enable one skilled in the art to utilize the present disclosure in various examples and with various modifications as are suited to the particular use contemplated. The features of the examples described herein may be combined in all possible combinations of methods, apparatus, modules, systems, and computer program products.

[0033] It is also noted herein that while the above describes examples, these descriptions should not be viewed in a limiting sense. Rather, there are several variations and modifications which may be made without departing from the scope as defined in the appended claims.