FactomProject / factomd-bench

tools for analyzing json-encoded factomd logs

factomd-bench

tools to inject json-encoded logs and produce reports

Loading data

• create tgz from logs after a simulation run tgz files should have a folder simTest/fnode00<variouslogs>.txt

#snippet from test_set2.sh in FD-798 branch of factomd

function stashlogs() {
  tar -czf $1.tgz simTest/*.txt
}

GIVEN_NODES=LF MAX_BLOCKS=40 BATCHES=10 ENTRIES=1000 DELAY_BLOCKS=1 DROP_RATE=100 ./test.sh
stashlogs test_LF_d100

• copy to a subolder like so

v5.1.0/
├── dataload.py
├── laptop
│   ├── extract.sh
│   ├── test_LLLLLF_d100.tgz
│   ├── test_LLLLLLLLF_d100.tgz
│   └── test_LLLLLLLLFF_d100.tgz
└── load.sh

• Extract some log files using extract.sh

v5.1.0/
├── dataload.py
├── laptop
│   ├── extract.sh
│   ├── test_LLLLLF_d100.tgz
│   ├── test_LLLLLFF_d100
│   │   ├── FNode0_missing_messages.txt
│   │   └── FNode0_simTest.txt
│   ├── test_LLLLLLLLF_d100.tgz
│   ├── test_LLLLLLLLFF_d100
│   │   ├── FNode0_missing_messages.txt
│   │   └── FNode0_simTest.txt
│   └── test_LLLLLLLLFF_d100.tgz
└── load.sh

• load them into the db using ./v5.1.0/load.sh or ./rebuild.sh to to run all

review log labels in db:

load=# select * from log_runs;
 id |         label          |             ts
----+------------------------+----------------------------
  1 | v5_test_LLLLLLLFF_d100 | 2019-01-31 17:07:33.116239
  2 | v5_test_LLLLLLLF_d100  | 2019-01-31 17:08:19.244705
  3 | v6_test_LLLLLLLFF_d100 | 2019-01-31 17:09:05.469478
  4 | v6_test_LLLLLLLF_d100  | 2019-01-31 17:09:52.358996
(4 rows)

^ this allows you to use WHERE run=X when comparing log data between runs