You are browsing documentation for a version other than the latest stable release. Switch to the latest stable release, 2.1.
This sections lists all the configuration options in
mender.conf. Some of
these options can also be modified using Yocto variables.
Specifies the location of the public key used to verify signed updates, and also enables signed-updates-only mode when it is set. If set the client will reject incorrectly signed updates, or updates without a signature. See also the section about signing and verification.
An integer that sets the number of seconds to wait between each attempt to download an update file. Note that the client may attempt more often initially to enable rapid upgrades, but will gradually fall back to this value if the server is busy. See also the section about polling intervals.
The Linux device that contains root filesystem A. This is set by the build system based on Yocto configuration and rarely needs to be modified.
The Linux device that contains root filesystem B. This is set by the build system based on Yocto configuration and rarely needs to be modified.
The server URL which is used as the basis for API requests. This should be set
to the server that runs the Mender server services. It should include the whole
https:// and a trailing slash.
The location of the public certificate of the server, if any. If this certificate is missing, or the one presented by the server doesn't match the one specified in this setting, the server certificate will be validated using standard certificate trust chains.
The interval for which the script is allowed to keep retrying. After this interval has expired, another retry attempt will be treated as a failure and if possible, the update will be rolled back.
The timeout before a state script that previously returned the special retry return code (21), will be run again.
The number of seconds to wait for any state script to terminate. If a script exceeds this running time, its process group will be killed and Mender will continue, treating the script as having failed. See also the section about state scripts.
A token which identifies which tenant a device belongs to. This is only relevant if using Hosted Mender.
The location where deployment logs will be written. This must be on a persistent partition to work correctly.
An integer that sets the number of seconds to wait between each check for a new update. Note that the client may occasionally check more often if there has been recent activity on the device. See also the section about polling intervals.