LAB INVENTORY VERSION 2

Requirements:

• Multiple sessions can read from the database simultaneously, but allow only one session at a time modify the database and lock reads while writing.
• Include platform profiles that contain the information about gathering statistics on memory, CPUs, FPGA, line card and cost.
• Verify that the host file and the testbeds are in sync
• Generate Reports: Lab wide summary, testbed wide summary, specific hardware (ie: all T1 cards), unsupported devices, user access, testbed additions/removals, hostfile <-> device mismatch, connectivity

Program Flow:

• lock file  (only one real time update at a time (uid & date to lock and log files))
• allow non destructive reads from: previous database & real time query
• source config files
• parse options
• read host file
• copy hostfile local
• parse
• verify/update against previous database (option specific)
• compare to old database
• update, copy, log new & removed testbeds
• generate a list of IPs and device types as specified by the host file
• probe requested devices
• verify connectivity (log failure reason: no connectivity, no passwd set, etc.)
• verify device type
• execute probe list appropriate for the device
• memory
• CPU count & versions
• Feature card count and versions
• Switch card count & versions
• line cards (versions?)
• Image version
• ASPs
• log off with error checking
• Report requested info to reports and/or stdout
• update database
• lock file for reading database
• no read access at this time
• move old database
• create new database
• remove read lock file
• read access restored
• remove write lock file (if a destructive run was initiated)
• write access restored

Before Coding:

• man page
• define directory structure
• set command line options
• create pseudo config file(s)
• list of libraries and their functions
• define a database format (would prefer some ASCII format)
• log file formats

Log Files:

• connectivity
• user access
• errors
• testbed additions
• testbed removals
• hostfile <-> device mismatch

Implementation Outline:

1. Develop a skeleton library for one device
1. include live mem read
2. include live cpu version
2. Write the skeleton main control function
1. write "probe requested devices"
2. write "Read host file"
3. write stdout display function
3. Develop lock file library
1. request to read
2. request to freeze writes
3. request to write
4. release request to write
4. Command line options
1. Develop the flow logic for the different options
2. Write a function to parse the command line and select the correct path
5. Develop format and interface functions for database
1. Add "Verify/Update against previous database" to the main control function
2. Add "Update database"