Query Progress Indicator
Many modern software systems provide progress indicators (PIs) for long-running tasks (e.g., file downloads and software installations). Typically, a progress indicator estimates how much of the task has been completed and when the task will finish. We are trying to develop progress indicators for database queries that continuously update the elapsed time, the estimated remaining time and percentage of completion for a given query. Ideally, a query progress indicator should satisfy the following goals: (i) it should provide estimates that accurate reflect the actual progress of the query, (ii) it should provide estimates at sufficiently fine granularity, (iii) it should take full advantage of information available during its execution, and (iv) it should impose minimal overhead on the actual execution of the query. In addition to estimating the progress for queries running on SQL Server, we are also working on progress indicators for PDW, Hadoop and Polybase systems.
- Jessie Li
- Rimma V. Nehme
- Jeff Naughton
- Jiexing Li, Rimma V. Nehme, Jeffrey Naughton: Toward Progress Indicators on Steroids for Big Data Systems, Conference on Innovative Data Systems Research (CIDR), 2013.
- Jiexing Li, Rimma V. Nehme, Jeffrey Naughton: GSLPI: a Cost-based Query Progress Indicator, International Conference on Data Engineering (ICDE), 2012.
- Jiexing Li, Arnd Christian König, Vivek R. Narasayya, Surajit Chaudhuri: Robust Estimation of Resource Consumption for SQL Queries using Statistical Techniques. PVLDB 5(11): 1555-1566 (2012)
- Toward Progress Indicators on Steroids for Big Data Systems. Conference on Innovative Data Systems Research (CIDR), 2013 (slides).
- GSLPI: a Cost-based Query Progress Indicator. International Conference on Data Engineering (ICDE), 2012 (slides).