Cloud replication based on adaptive quality of service
Disclosed herein are methods, systems, and processes to perform cloud replication based on adaptive Quality of Service. A replication stream is monitored over a period of time. The replication stream includes write operations issued by an application, and is associated with preset parameters. Replication parameters applicable to the replication stream are determined. The replication parameters are configured to be used in a replication operation. The preset parameters and the replication parameters are stored.
1. A method comprising:
measuring replication parameters of a replication stream, wherein the replication stream comprises write operations issued by a plurality of applications, respectively, and wherein the replication parameters comprise rates at which the plurality of applications, respectively, issue write operations;
aggregating data of the write operations into update sets corresponding to the plurality of applications, respectively;
determining a plurality of preset replication intervals assigned to the plurality of applications, respectively;
generating adjusted replication intervals for the plurality of applications, respectively, by increasing or decreasing replication intervals for the plurality of applications, respectively, based on changes in the rates at which the plurality of applications issue write operations;
calculating a first cost of replicating the aggregated data for a first application of the plurality of applications, to the target site via a network using a first adjusted replication interval for the first application;
calculating a second cost of replicating the aggregated data for the first application to the target site via the network using a first preset parameter of the plurality of preset parameters;
comparing the first and second costs;
if the first cost exceeds the second cost, replicating the aggregated data for the first application to the target site at the first preset replication interval;
if the second cost exceeds the first cost, replicating the aggregated data for the first application to the target site at the first adjusted replication interval.
2. The method of claim 1 wherein the first and second costs are measured in the same units of currency.
3. The method of claim 1 wherein the first and second costs are calculated using a cost of transmitting data via the network per unit of data.
4. The method of claim 1 wherein the first and second costs of transmitting the aggregated data via the network is calculated per unit of data.
5. The method of claim 1 , wherein
the replication stream is received from a plurality of virtual machines via I/O taps,
the plurality of virtual machines execute the plurality of applications, respectively;
each application of the plurality is assigned a workload priority of a plurality of workload priorities.
6. The method of claim 5 , wherein the preset replication intervals are based on the workload priorities, respectively, assigned to the plurality of applications, respectively.
7. A non-transitory computer readable storage medium comprising program instructions executable to:
measure replication parameters of a replication stream, wherein the replication stream comprises write operations issued by a plurality of applications, respectively, and wherein the replication parameters comprise rates at which the plurality of applications, respectively, issue write operations;
aggregate data of the write operations into update sets corresponding to the plurality of applications, respectively;
determine a plurality of preset replication intervals assigned to the plurality of applications, respectively;
generate adjusted replication intervals for the plurality of applications, respectively, by increasing or decreasing replication intervals for the plurality of applications, respectively, based on changes in the rates at which the plurality of applications issue write operations;
calculate a first cost of replicating the aggregated data for a first application of the plurality of applications, to the target site via a network using a first adjusted replication interval for the first application;
calculate a second cost of replicating the aggregated data for the first application to the target site via the network using a first preset parameter of the plurality of preset parameters;
compare the first and second costs;
if the first cost exceeds the second cost, replicate the aggregated data for the first application to the target site at the first preset replication interval;
if the second cost exceeds the first cost, replicate the aggregated data for the first application to the target site at the first adjusted replication interval.
8. The non-transitory computer readable storage medium of claim 7 , wherein the first and second costs are measured in the same units of currency.
9. The non-transitory computer readable storage medium of claim 7 , wherein the cost is calculated based on a cost of transmitting data via the network per unit of data.
10. The non-transitory computer readable storage medium of claim 7 , wherein the first and second costs are calculated using a cost of transmitting data via the network per unit of data.
11. 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:
measure replication parameters of a replication stream, wherein the replication stream comprises write operations issued by a plurality of applications, respectively, and wherein the replication parameters comprise rates at which the plurality of applications, respectively, issue write operations;
aggregate data of the write operations into update sets corresponding to the plurality of applications, respectively;
determine a plurality of preset replication intervals assigned to the plurality of applications, respectively;
generate adjusted replication intervals for the plurality of applications, respectively, by increasing or decreasing replication intervals for the plurality of applications, respectively, based on changes in the rates at which the plurality of applications issue write operations;
calculate a first cost of replicating the aggregated data for a first application of the plurality of applications, to the target site via a network using a first adjusted replication interval for the first application;
calculate a second cost of replicating the aggregated data for the first application to the target site via the network using a first preset parameter of the plurality of preset parameters;
compare the first and second costs;
if the first cost exceeds the second cost, replicate the aggregated data for the first application to the target site at the first preset replication interval;
if the second cost exceeds the first cost, replicate the aggregated data for the first application to the target site at the first adjusted replication interval.
12. The system of claim 11 , wherein the first and second costs are measured in the same units of currency.
13. The system of claim 11 , wherein the first and second costs are calculated using a cost of transmitting data via the network per unit of data.