Query performance problem determination is usually performed manually in consultation with experts through the analysis of query plans. However, this is an excessively time consuming, human error-prone, and costly process. GALO is a novel system that automates this process. The tool automatically learns recurring problem patterns in query plans over workloads in an offline learning phase to build a knowledge base of plan rewrite remedies. GALO's knowledge base is built on RDF and SPARQL, which is well-suited for manipulating and querying over SQL query plans, which are graphs themselves. It then uses the knowledge base online to re-optimize queries queued for execution to improve performance, often quite dramatically.

Query optimization is a hallmark of database systems. When an SQL query runs more expensively than is viable or warranted, determination of the performance issues is usually performed manually in consultation with experts through the analysis of query's execution plan (QEP). However, this is an excessively time consuming, human error-prone, and costly process. GALO is a novel system that automates this process. The tool automatically learns recurring problem patterns in query plans over workloads in an offline learning phase, to build a knowledge base of plan-rewrite remedies. It then uses the knowledge base online to re-optimize queries often quite drastically. GALO's knowledge base is built on RDF and SPARQL, W3C graph database standards, which is well suited for manipulating and querying over SQL query plans, which are graphs themselves. GALO acts as a third-tier of re-optimization, after query rewrite and cost-based optimization, as a query plan rewrite . For generality, the context of knowledge base problem patterns, including table and column names, is abstracted with canonical symbol labels. Since the knowledge base is not tied to the context of supplied QEPs, table and column names are matched automatically during the re-optimization phase. Thus, problem patterns learned over a particular query workload can be applied in other query workloads. GALO's knowledge base is also an invaluable tool for database experts to debug query performance issues by tracking to known issues and solutions as well as refining the optimizer with new tuned techniques by the development team. We demonstrate an experimental study of the effectiveness of our techniques over synthetic TPC-DS and real IBM client query workloads.

query performance problem determination is often performed by analyzing query execution plans (QEPs) in addition to other performance data. As the query workloads that organizations run have become larger and more complex, analyzing QEPs manually even by experts has become a very time consuming and cumbersome task. Most performance diagnostic tools help with identifying problematic queries and most query tuning tools address a limited number of known problems and recommendations. We present the OptImatch system that offers a way to (a) look for varied user defined problem patterns in QEPs and (b) automatically get recommendations from an expert provided and user customizable knowledge base. Existing approaches do not provide the ability to perform workload analysis with flexible user defined patterns, as they lack the ability to impose a proper structure on QEPs. We introduce a novel semantic web system that allows a relatively naive user to search for arbitrary patterns and to get solution recommendations stored in a knowledge base. Our methodology includes transforming a QEP into an RDF graph and transforming a GUI based user-defined pattern into a SPARQL query through handlers. The SPARQL query is matched against the abstracted RDF graph, and any matched portion of the abstracted RDF graph is relayed back to the user. With the knowledge base, the OptImatch system automatically scans and matches interesting stored patterns in a statistical way as appropriate and returns the corresponding recommendations. Although the knowledge base patterns and solution recommendations are not in the context of the user supplied QEPs, the context is adapted automatically through the handler tagging interface. We test the performance and scalability of our framework to demonstrate its efficiency using a real query workload. We also perform a user study to quantify the benefits of the approach in terms of precision and time compared to manually searching for patterns.

Query performance problem determination is usually performed by analyzing query execution plans (QEPs). Analyzing complex QEPs is excessively time consuming and existing automatic problem determination tools do not provide ability to perform analysis with flexible user-defined problem patterns. We present the novel OptImatch system that allows a relatively naive user to search for patterns in QEPs and get recommendations from an expert and user customizable knowledge base. Our system transforms a QEP into an RDF graph. We provide a web graphical interface for the user to describe a pattern that is transformed with handlers into a SPARQL query. The SPARQL query is matched against the abstracted RDF graph and any matched parts of the graph are relayed back to the user. With the knowledge base the system automatically matches stored patterns to the QEPs by adapting dynamic context through developed tagging language and ranks recommendations using statistical correlation analysis.