Overview

Architecture
The Diagnose project will use Django as the web framework, below is the Django architecture:

 

1.	The URL dispatcher (urls.py) maps the requested URL to a view function and calls it. If catching is enabled, the view function can check to see if a cached version of the page exists and bypass all further steps, returning the cached version, instead, Note that this page-level caching is only one available caching option in Django. You can cache more granularly, as well.
2.	The view function (usually in views.py) performs the requested action, which typically involves reading or writing to the database. It may include other tasks, as well.
3.	The model (usually in models.py) defines the data in Python and interacts with it, although typically contained in a relational database (MySQL in diagnose project), other data storage mechanisms are possible as well (XMl, text files, LDAP, etc.).
4.	After performing any requested tasks, the view returns an HTTP response object (usually after passing the data through a template) to the web browser. Optionally, the view can save a version of the HTTP response object in the caching system for a specified length of time.
5.	Templates typically return HTML pages. The Django template language offers HTML authors a simple-to-learn syntax while providing all the power needed for presentation logic.

Performance
This module will show the system performance statics
Critical resource display
Data structure: 
Thanks to SDM measurement table SDM_MEAS_LOAD_STATION_HSS as original data, below info are stored in DB:
Table: SDM_MEAS_LOAD_STATION_HSS
FEName	Day	Time	MsgClass	Blade	Duration	STATION_ID	ACTIVE_IN_AO_NUM	ROLLBACK_ACTIVE_JOBNUM	　	　
Note:
FEName -> NodeName
Remove MsgClass, Blade
UI
Time selection: duration or point
Resource selection: single or multiple resource 
Blade selection: single or multiple blades
Output: 
CPU usage per process
Data structure: 
Thanks to mCAS measurement table MS_PROCESS_MEAS as original data, below info are stored in DB:
Table: MS_PROCESS_MEAS
Node	Day	Time	Blade	Process	CPU
UI
Time selection: duration or point
Blade selection: single or multiple blades

Trouble shooting
This module will provide a simple troubleshooting feature:
Solution
Search PRID/PUID in Description and return all lines containing this sub info
Data structure: 
NodeName	Day	Time	MsgClass	Blade	Title	Description	Log ID
UI
User input: sub info(PUID/PRID)
Message type: MAR, SAR, UDR…
Timeslot: duration
Abnormal logs
This module will show the histogram of log ID stored in DB, such as top 10 error types.
Solution
Fetch all record in a period and Sort by LogID, then display it using histogram. 
UI
Timeslot: duration
Timeslot: duration
Number: digit

Utils
sdm_dmt_parser.py
This script is used to generate table schema
Input: /vob/sdm_mcas/src/utils/meas/SDM.dmt
Output: models.py

class SDM_MEAS_LOAD_STATION_HSS(models.Model):
    FEName = models.TextField(blank = True, null = True)
    Day = models.DateField()
    Time = models.TimeField()
    MsgClass = models.TextField(blank = True, null = True)
    Blade = models.TextField(blank = True, null = True)
    Duration = models.TextField(blank = True, null = True)
    STATION_ID = models.TextField(blank = True, null = True)
    PROCESS_NAME = models.TextField(blank = True, null = True)
    ACTIVE_IN_AO_NUM = models.TextField(blank = True, null = True)
    ACTIVE_IN_AO_NUM_MEAN = models.TextField(blank = True, null = True)

omlog_parser.py
This script is used to parse OMlog and IMSRTlog etc.
Input: OMlog and IMSRTLog files
Output: txt files

meas_parser.py
this script is used to parse meas logs.
Input: measlog* files
Output: txt files