README.md

Cassandra

Be sure that environment variable $DATA is defined.

[ ! -z "$DATA" ] && \
docker run -d \
--name beehive-cassandra \
--net beehive \
-v ${DATA}/cassandra/:/var/lib/cassandra/ \
--ulimit memlock=-1 \
--ulimit nofile=100000 \
--ulimit nproc=32768 \
--cap-add IPC_LOCK \
cassandra:3.2 -R

For simple testing without much data you can omit option "-v" above. Without "-v" Cassandra data is not stored persistently and data is lost when the container is removed. Add port mapping "-p 7000:7000" if the beehive server does not run on the same host as the cassandra database.

Installation instructions for Cassandra without Docker:

http://docs.datastax.com/en/cassandra/2.0/cassandra/install/installDeb_t.html

Notes

After initialization, cassandra will not contain any databases (It will not include the "waggle" database). However, after the database has been created and populated, it can be interactively explored using a "cqlsh" prompt via the following commands.

To directly connect to cassandra:

docker run -it --net beehive --rm cassandra:3.2 cqlsh beehive-cassandra

To view database (after it is created), e.g.:

use waggle;
DESCRIBE TABLES;
SELECT * FROM nodes;
SELECT * FROM sensor_data;

Troubleshooting

In case of multiple corrupt commit log files, files can be deleted in a more automated way using a bash loop. Start the cassandra container using --entrypoint=/bin/bash to get a terminal.

BAD_FILE='x'
while [ "${BAD_FILE}x" != 'x' ] ; do 
  export BAD_FILE=$(cassandra -f -R | grep CommitLogReplayException | grep -o "/var/lib/cassandra/commitlog/CommitLog-6-[0-9]*.log")
  echo "delete ${BAD_FILE}"
  rm -f ${BAD_FILE}
done