Read performance enhancement by enabling read from secondary in highly available cluster setup
A computer-implemented method for efficiently accessing a secondary storage in highly available clustered storage environment may include receiving a client-initiated data request at a secondary server coupled to a secondary storage; determining request information about the client-initiated data request; determining with the secondary server whether to process the client-initiated data request on the secondary storage based on the determined request information; and responsive to determining that the secondary server should process the client-initiated data request, processing the client-initiated data request by the secondary server to retrieve data from the secondary storage.
1. A system for efficiently accessing data in a highly available clustered storage, the system comprising:
a secondary server configured to receive, process, and distribute a client-initiated data request for accessing data;
a primary server coupled to the secondary server to receive the client-initiated data request and to provide data in response to the client-initiated data request;
a primary storage coupled to the primary server for providing data responsive to the client-initiated data request via the primary server; and
a secondary storage coupled to the secondary server for providing data responsive to the client-initiated data request via the secondary server, wherein, responsive to the received client-initiated data request, the secondary server is further configured to:
determine request information including:
a request input/output (I/O) type about the client-initiated data request;
a synchronization status of the primary storage and the secondary storage associated with the request; and
a location of the secondary storage associated with the request relative to the secondary server;
determine whether to process the client-initiated data request on the secondary server or on the primary server;
process the client-initiated data request using the secondary server to retrieve data from the secondary storage responsive to determining that:
the request I/O type of the client-initiated data request is a read request;
the primary storage and secondary storage associated with requested data are in synchronization; and
the location of the secondary storage associated with the requested data is within a determined distance to the secondary server; and
forward the client-initiated data request to the primary server to process the client-initiated data request responsive to determining that:
the request I/O type of the client-initiated data request is not a read request;
the primary storage and the secondary storage associated with the requested data are not in synchronization; or
the location of the secondary storage associated with the requested data is not within the determined distance to the secondary server.
2. The system of claim 1 , wherein:
the secondary server comprises a direct secondary interface module configured to retrieve requested data directly from the secondary storage.
3. The system of claim 2 , wherein the direct secondary interface module includes a request identification module to determine the request I/O type of the client-initiated data request.
4. The system of claim 2 , wherein the direct secondary interface module includes a locator module to determine the physical location of data corresponding to the client-initiated data request from a pool of storage devices including the primary storage and the secondary storage.
5. The system of claim 2 , wherein the direct secondary interface module includes a sync detector module to detect whether the primary storage and the secondary storage are in synchronization.
6. The system of claim 2 , wherein the direct secondary interface module includes a switch module for determining a working state to assign to the direct secondary interface module, the working state assigned based on a state of the secondary storage.
7. A computer-implemented method comprising:
receiving a client-initiated data request at a secondary server of a cluster of servers, the cluster of servers including a primary server coupled to a primary storage and the secondary server coupled to a secondary storage;
determining request information including:
a request input/output (I/O) type about the client-initiated data request;
a synchronization status of the primary storage and the secondary storage associated with the request and
a location of the secondary storage associated with the request relative to the secondary server;
determining with the secondary server whether to process the client-initiated data request on the secondary server or on the primary server;
processing the client-initiated data request by the secondary server to retrieve data from the secondary storage responsive to determining that:
the request I/O type of the client-initiated data request is a read request secondary server should process the client initiated data request;
the primary storage and secondary storage associated with requested data are in synchronization; and
the location of the secondary storage associated with the requested data is within a determined distance to the secondary server; and
forwarding, by the secondary server, the client-initiated data request to the primary server to process the client-initiated data request responsive to determining that:
the request I/O type of the client-initiated data request is not a read request;
the primary storage and the secondary storage associated with the requested data are not in synchronization; or
the location of the secondary storage associated with the requested data is not within the determined distance to the secondary server.
8. The computer-implemented method of claim 7 , wherein processing the client-initiated data request by the secondary server includes responding directly to the client-initiated data request by retrieving data from the secondary storage through a direct secondary interface.
9. The computer-implemented method of claim 8 , wherein:
the secondary server includes the direct secondary interface; and
the determined distance includes the direct secondary interface directly attaching the secondary server to the secondary storage.
10. The computer-implemented method of claim 7 , wherein:
the secondary server is clustered with the primary server in a highly available clustered storage system; and
the primary storage and the secondary storage are mirrored storage connected through a network.
11. The computer-implemented method of claim 7 , wherein determining with the secondary server whether to process the client-initiated data request on the secondary storage includes determining whether the secondary storage and the secondary server are on a same node.
12. The computer-implemented method of claim 7 , wherein determining with the secondary server whether to process the client-initiated data request on the secondary storage includes determining a working state to assign to a direct secondary interface between the secondary server and the secondary storage.
13. A system comprising:
a primary server comprising a first processor and a first memory;
a primary storage coupled to the primary server;
a secondary server comprising a second processor and a second memory; and
a secondary storage coupled to the secondary server,
wherein the secondary server further comprises:
means for receiving a client-initiated data request at the secondary server;
means for determining request information including:
a request input/output (I/O) type about the client-initiated data request;
a synchronization status of the primary storage and the secondary storage associated with the request and
a location of the secondary storage associated with the request relative to the secondary server; and
means for determining whether to process the client-initiated data request on the secondary server or on the primary server based on the determined request information, wherein:
processing the client-initiated data request by the secondary server to retrieve data from the secondary storage responsive to determining that:
the request I/O type of the client-initiated data request is a read request;
the primary storage and secondary storage associated with requested data are in synchronization; and
the location of the secondary storage associated with the requested data is within a determined distance to the secondary server; and
forwarding, by the secondary server, the client-initiated data request to the primary server to process the client-initiated data request responsive to determining that:
the request I/O type of the client-initiated data request is not a read request;
the primary storage and the secondary storage associated with the requested data are not in synchronization; or
the location of the secondary storage associated with the requested data is not within the determined distance to the secondary server.
14. The system of claim 13 , wherein processing the client-initiated data request by the secondary server includes responding directly to the client-initiated data request by retrieving data from the secondary storage through a direct secondary interface.
15. The system of claim 14 , wherein:
the secondary server includes the direct secondary interface; and
the determined distance includes the direct secondary interface directly attaching the secondary server to the secondary storage.
16. The system of claim 13 , wherein:
the secondary server is clustered with the primary server in a highly available clustered storage system; and
the primary storage and the secondary storage are mirrored storage connected through a network.
17. The system of claim 13 , wherein the means for determining with the secondary server whether to process the client-initiated data request on the secondary storage includes means for determining whether the secondary storage and the secondary server are on a same node.
18. The system of claim 13 , wherein the means for determining with the secondary server whether to process the client-initiated data request on the secondary storage includes means for determining a working state to assign to a direct secondary interface between the secondary server and the secondary storage.
19. The computer-implemented method of claim 7 , further comprising:
determining the determined distance using a physical distance between the secondary server and the primary storage.
20. The computer-implemented method of claim 7 , further comprising:
determining the determined distance using a latency in data transfer between the secondary server and the primary storage.