Application performance in replication environments
Disclosed herein are methods, systems, and processes to improve application performance in replication environments. A read command configured to cause a read operation to be performed on a unit of storage in a data volume is received. A state of a lock is determined. The state of the lock indicates whether a write operation is ongoing, and the write operation results in data being written to a storage replication log and to the data volume. The write operation is ongoing because writing the data to the storage replication log and to the data volume has not completed. In response to a determination that the write operation is not ongoing, the read operation is allowed to be performed on the unit of storage.
1. A computer-implemented method comprising:
receiving a read command, wherein
the read command is configured to cause a read operation to be performed on a unit of storage, and
the unit of storage is comprised in a data volume;
determining a state of a lock, wherein
the state of the lock indicates whether a write operation is ongoing,
the write operation results in
writing data to a location in a storage replication log corresponding to the unit of storage, and
writing the data to a location in the data volume corresponding to the unit of storage, and
a determination that the write operation is not ongoing is made only if the writing the data to the location in the storage replication log, and the writing the data to the location in the data volume, have both been completed; and
in response to a determination that the state of the lock indicates that the write operation is not ongoing, allowing the read operation to be performed on the unit of storage.
2. The computer-implemented method of claim 1 , further comprising:
assigning a current generation number to the data volume; and
receiving a first request to lock the data volume, wherein
the first request is received due to the write operation, and
the write operation is asynchronous.
3. The computer-implemented method of claim 2 , further comprising:
upon receiving the first request,
assigning the current generation number to the first request,
setting a synchronous write acknowledgement flag on the data volume,
resetting a current generation input/output count of the data volume to zero, and
incrementing the current generation number by one.
4. The computer-implemented method of claim 3 , further comprising:
switching the data volume from a first write acknowledgement mode to a second write acknowledgement mode; and
receiving a second request to lock the data volume to the read operation, wherein
the second request is received due to the write operation, and
the write operation is synchronous and ongoing by virtue of the data volume being switched to the second write acknowledgement mode.
5. The computer-implemented method of claim 4 , further comprising:
upon receiving the second request and switching the data volume to the second write acknowledgement mode,
assigning the current generation number to the second request,
setting a no lock flag on the data volume to indicate that the read operation does not require the lock,
resetting the current generation input/output count of the data volume to zero, and
incrementing the current generation number by one; and
setting an asynchronous write acknowledgement flag on the data volume.
6. The computer-implemented method of claim 5 , further comprising:
determining that the read operation or the write operation is associated with one or more computing devices operating in a cluster, wherein
the data volume is part of a plurality of data volumes, and
the plurality of data volumes are part of the cluster; and
setting a synchronous write acknowledgement flag on each of the plurality of data volumes;
resetting the current generation input/output count of the plurality of data volumes to zero;
incrementing the current generation number by one;
waiting for a previous generation input/output count to become zero; and
setting a no lock flag on each of the plurality of data volumes.
7. The computer-implemented method of claim 6 , further comprising:
setting the no lock flag on the plurality of data volumes;
resetting the current generation input/output count of the plurality of data volumes to zero;
incrementing the current generation number by one; and
setting the asynchronous write acknowledgement flag on the plurality of data volumes.
8. A non-transitory computer readable storage medium comprising program instructions executable to:
receive a read command, wherein
the read command is configured to cause a read operation to be performed on a unit of storage, and
the unit of storage is comprised in a data volume;
determine a state of a lock, wherein
the state of the lock indicates whether a write operation is ongoing, and
the write operation results in
writing data to a location in a storage replication log corresponding to the unit of storage, and
writing the data to a location in the data volume corresponding to the unit of storage, and
a determination that the write operation is not ongoing is made only if the writing the data to the location in the storage replication log, and the writing the data to the location in the data volume, have both been completed; and
in response to a determination that the state of the lock indicates that the write operation is not ongoing, allow the read operation to be performed on the unit of storage.
9. The non-transitory computer readable storage medium of claim 8 , further comprising:
assigning a current generation number to the data volume; and
receiving a first request to lock the data volume, wherein
the first request is received due to the write operation, and
the write operation is asynchronous.
10. The non-transitory computer readable storage medium of claim 9 , further comprising:
upon receiving the first request,
assigning the current generation number to the first request,
setting a synchronous write acknowledgement flag on the data volume,
resetting a current generation input/output count of the data volume to zero, and
incrementing the current generation number by one.
11. The non-transitory computer readable storage medium of claim 10 , further comprising:
switching the data volume from a first write acknowledgement mode to a second write acknowledgement mode; and
receiving a second request to lock the data volume to the read operation, wherein
the second request is received due to the write operation, and
the write operation is synchronous and ongoing by virtue of the data volume being switched to the second write acknowledgement mode.
12. The non-transitory computer readable storage medium of claim 11 , further comprising:
upon receiving the second request and switching the data volume to the second write acknowledgement mode,
assigning the current generation number to the second request,
setting a no lock flag on the data volume to indicate that the read operation does not require the lock,
resetting the current generation input/output count of the data volume to zero, and
incrementing the current generation number by one; and
setting an asynchronous write acknowledgement flag on the data volume.
13. The non-transitory computer readable storage medium of claim 12 , further comprising:
determining that the read operation or the write operation is associated with one or more computing devices operating in a cluster, wherein
the data volume is part of a plurality of data volumes, and
the plurality of data volumes are part of the cluster; and
setting a synchronous write acknowledgement flag on each of the plurality of data volumes;
resetting the current generation input/output count of the plurality of data volumes to zero;
incrementing the current generation number by one;
waiting for a previous generation input/output count to become zero; and
setting a no lock flag on each of the plurality of data volumes.
14. The non-transitory computer readable storage medium of claim 13 , further comprising:
setting the no lock flag on the plurality of data volumes;
resetting the current generation input/output count of the plurality of data volumes to zero;
incrementing the current generation number by one; and
setting the asynchronous write acknowledgement flag on the plurality of data volumes.
15. A system comprising:
one or more processors; and
a memory coupled to the one or more processors, wherein the memory stores program instructions executable by the one or more processors to
receive a read command, wherein
the read command is configured to cause a read operation to be performed on a unit of storage, and
the unit of storage is comprised in a data volume,
determine a state of a lock, wherein
the state of the lock indicates whether a write operation is ongoing, and
the write operation results in
writing data to a location in a storage replication log corresponding to the unit of storage, and
writing the data to a location in the data volume corresponding to the unit of storage, and
a determination that the write operation is not ongoing is made only if the writing the data to the location in the storage replication log, and the writing the data to the location in the data volume, have both been completed, and
in response to a determination that the state of the lock indicates that the write operation is not ongoing, allow the read operation to be performed on the unit of storage.
16. The system of claim 15 , further comprising:
assigning a current generation number to the data volume; and
receiving a first request to lock the data volume, wherein
the first request is received due to the write operation, and
the write operation is asynchronous.
17. The system of claim 16 , further comprising:
upon receiving the first request,
assigning the current generation number to the first request,
setting a synchronous write acknowledgement flag on the data volume,
resetting a current generation input/output count of the data volume to zero, and
incrementing the current generation number by one.
18. The system of claim 17 , further comprising:
switching the data volume from a first write acknowledgement mode to a second write acknowledgement mode;
receiving a second request to lock the data volume to the read operation, wherein
the second request is received due to the write operation, and
the write operation is synchronous and ongoing by virtue of the data volume being switched to the second write acknowledgement mode;
upon receiving the second request and switching the data volume to the second write acknowledgement mode,
assigning the current generation number to the second request,
setting a no lock flag on the data volume to indicate that the read operation does not require the lock,
resetting the current generation input/output count of the data volume to zero, and
incrementing the current generation number by one; and
setting an asynchronous write acknowledgement flag on the data volume.
19. The system of claim 18 , further comprising:
determining that the read operation or the write operation is associated with one or more computing devices operating in a cluster, wherein
the data volume is part of a plurality of data volumes, and
the plurality of data volumes are part of the cluster; and
setting a synchronous write acknowledgement flag on each of the plurality of data volumes;
resetting the current generation input/output count of the plurality of data volumes to zero;
incrementing the current generation number by one;
waiting for a previous generation input/output count to become zero; and
setting a no lock flag on each of the plurality of data volumes.
20. The system of claim 19 , further comprising:
setting the no lock flag on the plurality of data volumes;
resetting the current generation input/output count of the plurality of data volumes to zero;
incrementing the current generation number by one; and
setting the asynchronous write acknowledgement flag on the plurality of data volumes.