Agent v2 has a different philosophy and some flags are different. If you are considering upgrading, here is what you should keep in mind:
Version 2 brings a unified experience for both the Agent and the Client. As such,
blackfire-agent executable does not exist anymore. It has been
replaced by the
blackfire-agent command becomes
blackfire-agent --register command becomes
blackfire config command becomes
-v) cannot be used in the form
-v 4 (
space, integer). Instead, please use
-v, equal, integer), or
blackfire CLI by passing options to
stdin, you can pass directly the parameters via the command line.
"foo\nbar\n" | blackfire config becomes
client:config --client-id=foo --client-token=bar.
The output on
stderr has changed. If you used to pipe the output of
blackfire-agent, you might have to upgrade your scripts.
For example, the report of
blackfire curl was previously written on
stdout, it is now written on
We provide a new
2 tag for the
blackfire/blackfire Docker image that
points to the lastest version of the image.
There is also a
1 tag alongside the existing
latest tag that points to
the previous version of the image for backward compatibility.
Starting from Sept 1st (end of support of version 1), the
will follow version
2, and the
1 tag will be frozen.
Note that the default port of version
2 of the Docker image uses the
8307 port instead of
We introduced this change to be consistent with the default port of the
We have dropped the
/etc/init.d/blackfire-agent sysvinit script in favor of a
systemd service. The systemd service name stays
blackfire-agent. Use these
commands to manipulate the service:
sudo systemctl stop blackfire-agent to stop the service.
sudo systemctl start blackfire-agent to start the service.
sudo systemctl restart blackfire-agent to restart the service.
We have changed the default Blackfire installation location. It was previously
C:\Program Files\Blackfire and it is now installed in