Download presentation
Presentation is loading. Please wait.
Published byNora Cox Modified over 8 years ago
1
Behavior Isolation in Enterprise Systems Mohamed Mansour mansour@cc.gatech.edu
2
Feb14, 20072 Client 1 Message queue Travel Industry Example Client 2 Client 3 clearinghouse Airlines GDS
3
Feb14, 20073 GDS Scale Mission critical environment 24/7 11.5 million queries/days 2-16 seconds processing time ~10GB data set, 20% annual increase 8 updates per day, moving to seamless updates Message queue GDS
4
Feb14, 20074 Effect of Request Stream
5
Feb14, 20075 Why We Care? Business Consumer Loyalty Violates contractual agreements Technical Occurs even in highly engineered systems Can cause ripple effects
6
Feb14, 20076 Lets Just Fix it! Difficult to identify root cause Constant data changes Request stream dependency Sometimes can’t fix root cause 3 rd part libraries Interactions with OS, and H/W caches Complex code base
7
Feb14, 20077 I(solation) Queue Dynamic management of message streams Correlate message sequences with server behavior Learning phase Isolate undesired sequences Control phase Evaluation metrics Quality of Information metrics (QoI)
8
Feb14, 20078 Learning Phase Use online learning methods Statistical correlation [ICSOC 06] HMM [GIT-CERCS-06-11] Behavior Model Associate undesired behaviors with certain input patterns
9
Feb14, 20079 Control Phase Observe input message sequence Control sequence dispatched to each server to maintain QoI Dispatcher Reordering messages in queue
10
Feb14, 200710 I-Queue Applied to Worldspan Pricing Engine Affects customer relations Possible impact on consumer experience – less options Objective: return maximum number of alternate fares Problem Variable number of alternate fares for same query Root cause unknown
11
Feb14, 200711 Establishing Behavior Model Heuristics point to query geographies Geography based on From/To city pair, e.g. East Coast to EU Fare data stored in disk files separated by geography Use geo-locality as our predictor Goal: improve geo-locality
12
Feb14, 200712 Modified Queue Dispatcher Dispatcher maintains server execution history Request routed to an available server with matching geography Message queue GDS
13
Feb14, 200713 Evaluation Used real traces from Worldspan Set of about 1800 requests 20% process in 16 seconds Geography extracted from messages Hand-coded mapping from city pairs to geography code Processing times measured using Worldspan servers Completely static environment Simulations to measure geo-matching Compare different isolation points
14
Feb14, 200714 Improvement in Geo-locality Matching improves 6 times for min. farm size Matching can improve further by adding more servers
15
Feb14, 200715 Choosing the Right Metrics to Monitor Min. of 28 servers to avoid queuing delays Geo-match increases with more servers Queuing delay is not the best metric to monitor
16
Feb14, 200716 Future Directions
Similar presentations
© 2024 SlidePlayer.com Inc.
All rights reserved.