Now that you have imported some traces (as instructed in OllyDbgImport ), change directory (cd) into the tracectory folder (containing the src/ directory). From there, start the GUI with the command

python src/analysis/guiserver.py

This command will spawn a server listening on port 8080 on the local interface. Navigate to http://localhost:8080/ to see the UI. The GUI will show available databases under the "Open trace" menu.

Acceleration (optional)

The default graphing/zooming method based on querying the MongoDB database is quite slow, which is why an accelerated version is in the works. It uses a specific treeserver process that builds range trees of the time-address points to enable quick querying (with the downside of requiring a few GBs of memory). To use this faster version, you have change to the src/datastruct directory and

• export read and write access data to specific files (replace DBNAME with the name you used when creating the database using ventilator.py)

#!bash

./mongoexport.py DBNAME reads > DBNAME_reads
./mongoexport.py DBNAME writes > DBNAME_writes

#!bash
g++ -O3 rangetree.cpp treeserver.cpp -lzmq -o treeserver

#!bash
./treeserver 5665 *reads *writes

Example:

#!bash
./mongoexport.py t206 reads > t206_reads
./mongoexport.py t206 writes > t206_writes
g++ -O3 rangetree.cpp treeserver.cpp -lzmq -o treeserver
./treeserver 5665 *reads *writes