Add section to README on reboots
This commit is contained in:
1 parent
e51f89268c
commit
add13ff857
1 file changed
+22
@@ -437,6 +437,28 @@ Any running application containers will permanently lose connectivity
|
||||
with the weave network and have to be restarted in order to
|
||||
re-connect.
|
||||
|
||||
### Reboots
|
||||
|
||||
If the host you are running your containers on gets rebooted, the
|
||||
default behaviour of Docker is to restart any containers that were
|
||||
running, but weave relies on a bridge and virtual ethernet devices
|
||||
being set up outside of the containers, so the weave network will not
|
||||
function in this state.
|
||||
|
||||
For the weave router, the best thing is to stop any container that
|
||||
Docker may have restarted and re-launch, i.e. something like:
|
||||
|
||||
host1# weave stop
|
||||
host1# weave launch 10.0.0.1/16
|
||||
|
||||
After that, other containers may be re-attached if you can identify
|
||||
them, like this:
|
||||
|
||||
host1# weave attach 10.0.1.1/24 <mycontainer>
|
||||
|
||||
Alternatively, you may prefer to disable the auto-restart feature in
|
||||
Docker as described [here](https://docs.docker.com/articles/host_integration/).
|
||||
|
||||
## Installation with Boot2Docker
|
||||
|
||||
If you are running Docker inside the Boot2Docker VM, e.g. because you
|
||||
|
||||
Reference in new issue
Block a user