Data processing lock signal transmission
In accordance with one aspect of the present description, a node of the distributed computing system has multiple communication paths to a data processing resource lock which controls access to shared resources, for example. In this manner, at least one redundant communication path is provided between a node and a data processing resource lock to facilitate reliable transmission of data processing resource lock signals between the node and the data processing resource lock. Other features and aspects may be realized, depending upon the particular application.
1. A computer implemented method, comprising:
under control of a first data processing unit,
transmitting by a first data processing unit over a first communication path to a data processing lock device, a first request to be granted a data processing lock,
prior to determining that a communication failure has occurred over the first communication path in connection with the transmission of the first request; transmitting by said first data processing unit over a second communication path to the data processing lock device, a second request to be granted the data processing lock; and
determining whether the data processing lock was obtained by the first data processing unit in response to either of said transmissions of the first and second requests to be granted the data processing lock; and
determining whether communication failures occurred over both said first and second communication paths in connection with said transmissions of the first and second requests, respectively.
2. The method of claim 1 wherein said determining whether communication failures occurred over both said first and second communication paths includes initiating a timeout period to wait for a grant of a requested data processing lock, and determining whether the timeout period has expired prior to receiving a grant of either of the first and second requests to be granted a data processing lock.
3. The method of claim 2 wherein the determining whether the data processing lock was obtained includes determining prior to expiration of the timeout period, whether the data processing lock was obtained by the first data processing unit in response to either of said transmissions of the first and second requests to be granted the data processing lock.
4. The method of claim 1 wherein the data processing lock is a data unit lock which controls access to a shared unit of data.
5. The method of claim 1 wherein the data processing lock is a data processing resource lock which controls access to a shared hardware resource.
6. The method of claim 1 further comprising upon determining that the data processing lock was obtained by the first data processing unit in response to either of said transmissions of the first and second requests to be granted the data processing lock, performing by the first data processing unit a failover which includes accessing shared data processing resources using the data processing lock held by the first data processing unit.
7. The method of claim 6 wherein the first data processing unit is a server node of a storage controller.
8. The method of claim 1 further comprising:
under control of the first data processing unit,
transmitting by the first data processing unit over the first communication path to the data processing lock device, a first release of a data processing lock granted to the first data processing unit;
prior to determining that a communication failure has occurred over the first communication path in connection with the transmission of the first release, transmitting by said first data processing unit over the second communication path, a second release the data processing lock granted to first data processing unit; and
determining whether the data processing lock was released by the first data processing unit in response to either of said transmissions of the first and second releases of the data processing lock; and
determining whether communication failures occurred over both said first and second communication paths in connection with said transmissions of the first and second releases, respectively.
9. A computer implemented method, comprising:
under control of a first server node of a distributed computer storage system,
transmitting over a first communication path to a shared hardware resource lock device, a first request to be granted a shared hardware resource lock of the hardware resource lock device;
determining whether a communication failure has occurred over the first communication path in connection with the transmission of the first request; and
upon determining that a communication failure has occurred over the first communication path in connection with the transmission of the first request, transmitting over a second communication path to the shared hardware resource lock device, a second request to be granted the shared hardware resource lock of the shared hardware resource lock device.
10. The method of claim 9 further comprising receiving by said server node a grant of a request to be granted the shared hardware resource lock of the shared hardware resource lock device, and, performing by the server node a failover which includes accessing shared hardware resources using the shared hardware resource lock held by the server node.