Computational reuse in Networking with stream processing by using apache flink snapshot
Advisor Information
Spyridon Mastorakis
Location
MBSC Omaha Room 304 - G
Presentation Type
Oral Presentation
Start Date
4-3-2022 9:00 AM
End Date
4-3-2022 10:15 AM
Abstract
Reuse is used when there is a need to use an event that had already completed in the previous state. Computational reuse needs to be used in networking to get a past operation in any future event with low latency. To reuse any previous operation, the previous state should be saved anywhere as backup. Snapshot was used to save the previous state so that it can reuse the state if the same state is found in future operation. Snapshot is a mechanism that is provided by an open source framework named by apache flink. Apache flink is used to process high volume and velocity of data. Data can be analyzed in real time by using stream processing compared to batch processing. In this work, we used apache flink snapshot to store the previous state, reuse it if the same state appears in the future cases. Two independent events were considered where it can identify whether there was any similar state in the running events compared to previous events. This provides a computational efficiency not to use the similar state again.
Scheduling Link
1
Computational reuse in Networking with stream processing by using apache flink snapshot
MBSC Omaha Room 304 - G
Reuse is used when there is a need to use an event that had already completed in the previous state. Computational reuse needs to be used in networking to get a past operation in any future event with low latency. To reuse any previous operation, the previous state should be saved anywhere as backup. Snapshot was used to save the previous state so that it can reuse the state if the same state is found in future operation. Snapshot is a mechanism that is provided by an open source framework named by apache flink. Apache flink is used to process high volume and velocity of data. Data can be analyzed in real time by using stream processing compared to batch processing. In this work, we used apache flink snapshot to store the previous state, reuse it if the same state appears in the future cases. Two independent events were considered where it can identify whether there was any similar state in the running events compared to previous events. This provides a computational efficiency not to use the similar state again.