https://devzone.nordicsemi.com/f/nordic-q-a/85818/nrf5-sdk-for-mesh-v4-1-0-flash-manager-corrupt-page-due-to-power-cycleHi, I have been trying to chase down an issue with the flash manager occasionally raising an assertion in mesh_config_backend_init() after a power cycle of the nRF52840 we are using in our product. Occasionally, when we re-provision the radio and thenen-USTelligent Community 13Tue, 23 Jul 2024 12:44:09 GMThttps://devzone.nordicsemi.com/thread/495214?ContentTypeID=1Tue, 23 Jul 2024 12:44:09 GMT137ad170-7792-4731-bb38-c0d22fbe4515:f9018942-9503-4f78-aaa3-800c5b2f7a65Aniket_Shet<p>Hi Mttrinh,</p> <p>Could you please mention what exactly was the issue and what was the fix put?</p><div style="clear:both;"></div>https://devzone.nordicsemi.com/thread/359981?ContentTypeID=1Fri, 25 Mar 2022 11:49:17 GMT137ad170-7792-4731-bb38-c0d22fbe4515:0d7f4620-900a-4849-912d-79bc0f0ef596Mttrinh<p>Hi,</p> <p>After having our developer having a look at the issue it seems like something similar was reported for v4.1.0. The issue was fixed in v4.2.0. It is highly recommended to use our latest version(v5.0.0) as a lot of issues have been fix since version v4.1.0</p><div style="clear:both;"></div>https://devzone.nordicsemi.com/thread/358466?ContentTypeID=1Wed, 16 Mar 2022 14:48:09 GMT137ad170-7792-4731-bb38-c0d22fbe4515:7c63f35d-e73b-4c95-93d9-4d7a8f17dc16Peter Maxwell Warasila<p>Specifically we are seeing the assertion in `flash_manager_internal.h :: get_next_entry()`<br /><img alt=" " src="https://devzone.nordicsemi.com/resized-image/__size/320x240/__key/communityserver-discussions-components-files/4/pastedimage1647441626344v1.png" /></p> <p>The device roams between different meshes based on its location. I&#39;d say the design goal is once a day over a ten year life span, but realistically it should only happen on a weekly basis.</p> <p>At this point in testing to reproduce the error, I reprovision the device and then power cycle the radio by resetting the host (and therefore cutting power to the nRF52840) within a second of the configuration/provisioning being sent to the host. It is not reproduceable every time, but happens at least once every 50 cycles, closer to once every ten. I assume this has to do with the queued flash manager write back action mechanism, but I&#39;m not certain.</p> <p>I have also verified and seen this effect over dozens of nRF52840s, but I have not been debugging each time and cannot point to this particular flash page as part of the issue.</p> <p>I should clarify that we are using the nRF52840 as a mesh radio with the nRF5 SDK for Mesh Serial Interface module connected to a host processor over UART.</p><div style="clear:both;"></div>https://devzone.nordicsemi.com/thread/358453?ContentTypeID=1Wed, 16 Mar 2022 14:17:54 GMT137ad170-7792-4731-bb38-c0d22fbe4515:0b7976c0-fe74-4e15-99ef-88a66ee94264Mttrinh<p>Hi,</p> <p>What kind of assert are you getting from&nbsp;<span>mesh_config_backend_init()?</span></p> [quote user=""]We are definitely reprovisioning more often in the current product, so that may be contributing to this being more likely to occur.[/quote] <p>How often are you reprovisioning?</p><div style="clear:both;"></div>