Redundant distributed data storage system
An improved information management system is provided herein for managing secondary storage computing devices when one or more become unavailable. For example, a first secondary storage computing device in a pool of secondary storage computing devices can run a storage manager virtual machine (VM) that manages secondary copy and restore operations. A second secondary storage computing device in the pool can run a virtualization engine that manages the storage manager VM. The virtualization engine can replicate the storage manager VM data to other secondary storage computing devices in the pool. If the first secondary storage computing devices becomes unavailable, the second secondary storage computing device can instruct a third secondary storage computing device in the pool to launch a replicated storage manager VM using the replicated storage manager VM data stored therein. Thus, the third secondary storage computing device can act as the now unavailable first secondary storage computing device.
1. A networked information management system configured to automatically re-configure data storage computers due to hardware failures, the networked information management system comprising:
a first data storage computer comprising a first drive of a first type and a first drive of a second type, wherein a storage manager virtual machine (VM) manages one or more secondary copy operations in the networked information management system and runs in the first drive of the first type, wherein storage manager VM data is stored in the first drive of the first type, and wherein first data managed by the first data storage computer is stored in the first drive of the second type;
a second data storage computer comprising a second drive of the first type and a second drive of the second type, wherein first replicated storage manager VM data is stored in the second drive of the first type, and wherein second data managed by the second data storage computer is stored in the second drive of the second type; and
a third data storage computer comprising a third drive of the first type and a third drive of the second type, wherein a virtualization engine runs in the third drive of the first type, wherein second replicated storage manager VM data is stored in the third drive of the first type, wherein third data managed by the third data storage computer is stored in the third drive of the second type, and wherein the third data storage computer is configured to:
detect that the first data storage computer is unavailable,
instruct the second data storage computer to launch a replicated storage manager VM using the first replicated storage manager VM data, and
process a message received from the second data storage computer indicating that the storage manager VM is now running on the second data storage computer instead of the first data storage computer.
2. The networked information management system of claim 1 , wherein the third data storage computer is further configured to:
instruct the first data storage computer to replicate the storage manager VM data to form the first replicated storage manager VM data;
obtain the first replicated storage manager VM data; and
transmit the first replicated storage manager VM data to the second data storage computer for storage.
3. The networked information management system of claim 1 , wherein the third data storage computer is further configured to instruct the first data storage computer to replicate the storage manager VM data to form the first replicated storage manager VM data and transmit the first replicated storage manager VM data to the second data storage computer for storage.
4. The networked information management system of claim 1 , wherein the first drive of the second type, the second drive of the second type, and the third drive of the second type collectively form a secondary storage device.
5. The networked information management system of claim 1 , wherein replicated second data and replicated third data are stored in the first drive of the second type, wherein replicated first data and the replicated third data are stored in the second drive of the second type, and wherein the replicated first data and the replicated second data are stored in the third drive of the second type.
6. The networked information management system of claim 1 , wherein the first drive of the first type, the second drive of the first type, and the third drive of the first type collectively form a solid-state drive (SSD) storage pool.
7. The networked information management system of claim 1 , wherein first replicated virtualization engine data is stored in the first drive of the first type, and wherein second replicated virtualization engine data is stored in the second drive of the first type.
8. The networked information management system of claim 1 , wherein the second data storage computer is further configured to instruct a fourth data storage computer to launch a replicated virtualization engine in response to a determination that the third data storage computer is unavailable.
9. A computer-implemented method for automatically re-configuring data storage computers due to hardware failures, the computer-implemented method comprising:
detecting, by a first data storage computer comprising a first drive of a first type and a first drive of a second type, that a second data storage computer is unavailable, wherein a virtualization engine runs in the first drive of the first type, wherein first data managed by the first data storage computer is stored in the first drive of the second type, wherein the second data storage computer comprises a second drive of the first type and a second drive of the second type, wherein a storage manager virtual machine (VM) manages one or more secondary copy operations in a networked information management system that comprises the first data storage computer and runs in the second drive of the first type, wherein storage manager VM data is stored in the second drive of the first type, and wherein second data managed by the second data storage computer is stored in the second drive of the second type;
instructing, by the first data storage computer, a third data storage computer to launch a replicated storage manager VM using replicated storage manager VM data stored in a third drive of the first type of the third data storage computer, wherein the third data storage computer comprises the third drive of the first type and a third drive of the second type, and wherein third data managed by the third data storage computer is stored in the third drive of the second type; and
processing, by the first data storage computer, a message received from the third data storage computer indicating that the storage manager VM is now running on the third data storage computer instead of the second data storage computer.
10. The computer-implemented method of claim 9 , further comprising:
instructing the second data storage computer to replicate the storage manager VM data to form the replicated storage manager VM data;
obtaining the replicated storage manager VM data; and
transmitting the replicated storage manager VM data to the third data storage computer for storage.
11. The computer-implemented method of claim 9 , further comprising instructing the second data storage computer to replicate the storage manager VM data to form the replicated storage manager VM data and transmit the replicated storage manager VM data to the third data storage computer for storage.
12. The computer-implemented method of claim 9 , wherein the first drive of the second type, the second drive of the second type, and the third drive of the second type collectively form a secondary storage device.
13. The computer-implemented method of claim 9 , wherein replicated second data and replicated third data are stored in the first drive of the second type, wherein replicated first data and the replicated third data are stored in the second drive of the second type, and wherein the replicated first data and the replicated second data are stored in the third drive of the second type.
14. The computer-implemented method of claim 9 , wherein the first drive of the first type, the second drive of the first type, and the third drive of the first type collectively form a solid-state drive (SSD) storage pool.
15. The computer-implemented method of claim 9 , wherein first replicated virtualization engine data is stored in the second drive of the first type, and wherein second replicated virtualization engine data is stored in the third drive of the first type.
16. A non-transitory computer-readable medium storing instructions, which when executed by a first data storage computer comprising a first drive of a first type and a first drive of a second type, cause the first data storage computer to perform a method comprising:
detecting that a second data storage computer is unavailable, wherein a virtualization engine runs in the first drive of the first type, wherein first data managed by the first data storage computer is stored in the first drive of the second type, wherein the second data storage computer comprises a second drive of the first type and a second drive of the second type, wherein a storage manager virtual machine (VM) manages one or more secondary copy operations in a networked information management system that comprises the first and second data storage computers and runs in the second drive of the first type, wherein storage manager VM data is stored in the second drive of the first type, and wherein second data managed by the second data storage computer is stored in the second drive of the second type;
instructing a third data storage computer to launch a replicated storage manager VM using replicated storage manager VM data stored in a third drive of the first type of the third data storage computer, wherein the third data storage computer comprises the third drive of the first type and a third drive of the second type, and wherein third data managed by the third data storage computer is stored in the third drive of the second type; and
processing a message received from the third data storage computer indicating that the storage manager VM is now running on the third data storage computer instead of the second data storage computer.
17. The non-transitory computer-readable medium of claim 16 , wherein the method further comprises:
instructing the second data storage computer to replicate the storage manager VM data to form the replicated storage manager VM data;
obtaining the replicated storage manager VM data; and
transmitting the replicated storage manager VM data to the third data storage computer for storage.
18. The non-transitory computer-readable medium of claim 16 , wherein the method further comprises instructing the second data storage computer to replicate the storage manager VM data to form the replicated storage manager VM data and transmit the replicated storage manager VM data to the third data storage computer for storage.
19. The non-transitory computer-readable medium of claim 16 , wherein the first drive of the second type, the second drive of the second type, and the third drive of the second type collectively form a secondary storage device.
20. A networked information management system configured to automatically re-configure data storage computers due to hardware failures, the networked information management system comprising:
a first data storage computer comprising first physical hardware, wherein a storage manager virtual machine (VM) manages one or more secondary copy operations in the networked information management system and runs in the first data storage computer, and wherein the first data storage computer stores storage manager VM data;
a second data storage computer comprising second physical hardware, wherein the second data storage computer stores first replicated storage manager VM data; and
a third data storage computer comprising third physical hardware, wherein a virtualization engine runs in the third data storage computer, wherein the third data storage computer stores second replicated storage manager VM data, and wherein the third data storage computer is configured to:
detect that the first data storage computer is unavailable,
instruct the second data storage computer to launch a replicated storage manager VM using the first replicated storage manager VM data, and
process a message received from the second data storage computer indicating that the storage manager VM is now running on the second data storage computer instead of the first data storage computer.