Running a custom build

Once you built a custom version of ArangoDB (see Compiling), you may want to run it using existing data or possibly in isolation from an existing installation.

We assumes that you are in the root directory of the ArangoDB distribution and compiling has successfully finished.

Note that this guide is for Linux only.

Running in isolation

This part shows how to run your custom build with an empty database directory

# create data directory
mkdir /tmp/arangodb

# run
bin/arangod \
    --configuration etc/relative/arangod.conf\
     --database.directory /tmp/arangodb

Running with data

This part shows how to run your custom build with the config and data from a pre-existing stable installation.

ArangoDB’s developers may change the db file format and after running with a changed file format, there may be no way back. Alternatively you can run your build in isolation and dump and restore the data from the stable to your custom build.

When running like this, you must run the db as the arangod user (the default installed by the package) in order to have write access to the log, database directory etc. Running as root will likely mess up the file permissions - good luck fixing that!

# become root first
su

# now switch to arangod and run
su - arangod
bin/arangod --configuration /etc/arangodb/arangod.conf

❮ Compile on Windows Recompiling jemalloc ❯