OTA by HTTP Server
Posted: Tue Jun 09, 2020 9:46 pmMy sketch using OTP HTTP is approaching 50% so I am looking for ways to minimize memory capacity problems. The latest documentation has a paragraph that may relate to this issue but its meaning is completely opaque:
By default, OTA filesystem updates overwrite the target flash directly. This can lead to the file system being corrupted if there is a power outage during the update process. In order to use the same two step process that is used for OTA application firmware updates, set the ATOMIC_FS_UPDATE flag. Note that you will need to have enough unused space for the new filesystem image to be stored, hence is why this is not the default behaviour.
What does ATOMIC_FS_UPDATE do? What is the "two-step" process and what process is implemented instead of it? If the sketch approaches 50% size, will ATOMIC_FS_UPDATE help or hurt?
By default, OTA filesystem updates overwrite the target flash directly. This can lead to the file system being corrupted if there is a power outage during the update process. In order to use the same two step process that is used for OTA application firmware updates, set the ATOMIC_FS_UPDATE flag. Note that you will need to have enough unused space for the new filesystem image to be stored, hence is why this is not the default behaviour.
What does ATOMIC_FS_UPDATE do? What is the "two-step" process and what process is implemented instead of it? If the sketch approaches 50% size, will ATOMIC_FS_UPDATE help or hurt?