The methodology of system optimization

Background:

1. Complex business
2. Wide range of interfaces
3. High degree of concurrency
4. Large number of visits

Focus Point:

1. Performance Index
2. The association between modules
3. Combination of knowledge and methodology

3 Types of performance problems:

1. Performance failure is occurring
2. The trouble has happened, and it's automatically good.
3. The fault hasn't happened yet, but on the way to happening

Overall optimization ideas:

2. The important premise:
1. Understand the characteristics of the business
2. Master Baseline Information
3. Ensure ordered tuning:
1. First platform after application
2. First whole after local
3. Dynamic re-static first
4. See Time consumption distributions:
1. Overall distribution of response time
2. The analysis time is consumed in which modules, applications, and SQL.
6. Application Optimization Five steps:
2. System Research
1. Business background
1. Industry in which
2. Use the crowd
3. Load conditions
1. Number of online users
2. Peak hours
2. Technical framework:
1. Software and Hardware Checklist:
1. Application Software Deployment Checklist:
1. Front desk
2. Background
3. Data
2. Hardware Platform Checklist:
1. Host
2. Internet
3. Storage
2. Schema topology diagram
3. Interface Condition Description
3. System type
1. Type of Business
1. Production-Type 7x24
2. Auxiliary type can stop
2. Type of transaction
1. Oltp
2. Olap
3. OLTP and OLAP mixed
4. Master Baseline Information
1. Database
2. Host
5. Understand the state of the system
1. The fault is in progress
2. The fault has occurred, and the automatic reply
3. The fault will come
4. Information collection
1. Deploy third-party monitoring tools:
1. Ffcs_itsm
2. Dt
3. I3
4. NMON
2. System comes with tools:
1. Platform Information
1. Dynamic
2. Static
2. AWR, ADDM, ASH
3. Jvm
4. Various logs
3. Write your own script tool:
6. Problem Locator:
1. Positioning means:
1. System Resource situation
2. Response Time Mode:
1. Time distribution of the entire life cycle of a single transaction
2. The time distribution of the overall transaction of the whole system
2. Location Purpose:
1. Internal system problem or external interface problem
2. The whole problem or the local problem
1. Overall issues
1. Overall database issues
2. Front and back applications
2. Local issues
1. A module is slow
2. A menu is slow
3. Resource issues or application issues:
1. Resource issues
1. Insufficient platform hardware
2. Other programs run out of resources
2. Application issues
1. Which application area to drop in
2. Which interface module to drop
3. Which specific method or SQL to fall into
8. Analytical Solutions
1. Resource Utilization Analysis:
1. Analysis:
1. Insufficient CPU Memory IO
2. Network card traffic is too large
3. Storage failure
2. Solve:
1. Scaling strategy:
1. Replace Server, Store
2. Increase CPU memory, etc.
2. Resource reorganization:
1. Remove unrelated apps
2. Adjust resource allocation
1. According to the full-day pressure situation, staggered running resources
2. Depending on how busy the machine is, it is reasonable to allocate different applications on different hosts.
2. Application Analysis:
1. Application Analytics Solutions
1. Java
1. GC Hangs too large
2. WebLogic parameter configuration is incorrect
2. C + +
2. Database Analysis Solution:
1. Code issues:
1. Sql
2. Plsql
2. Model Design Issues
1. Design of the table
2. Design of the Index
3. Paradigm
3. Parameter issues
3. Schema Analysis Solution:
1. Cache
2. Asynchronous
3. Parallel
4. Distribution
5. Focus
10. Summary tracking:
1. Summarize:
1. This tuning report
2. This implementation of various documents
3. Lessons learned:
1. Lack of pre-launch performance stress testing:
1. LoadRunner
2. Ffcs Test Tool
2. Improper operation:
1. Not patched according to specifications
2. Feature upgrade introduces Bugs
3. Insufficient forecast of future business development
4. Lack of skills in development and maintenance personnel
2. Tracking:
1. Daily Patrol
2. Trend Analysis
3. Related training
1. Summary of this tuning case
2. Similar case sharing

The methodology of system optimization