https://devzone.nordicsemi.com/f/nordic-q-a/90994/nrf9160-nvs-write-flush-ncs-1-9-1-mfw-1-3-2Hello, My application stores settings in Zephyr NVS. During device production a python script running on PC sends RTT commands to write initial settings to NVS/flash. The setup ran flawlessly on my table, but factory had some issues as the settingsen-USTelligent Community 13Wed, 17 Aug 2022 07:29:43 GMThttps://devzone.nordicsemi.com/thread/381897?ContentTypeID=1Wed, 17 Aug 2022 07:29:43 GMT137ad170-7792-4731-bb38-c0d22fbe4515:9e02b397-261e-49fc-920a-800681980341plotkin1996<p>Thank you. That&#39;s probably an error in the production script or the way I implemented communication protocol than, will double-check.</p><div style="clear:both;"></div>https://devzone.nordicsemi.com/thread/381837?ContentTypeID=1Tue, 16 Aug 2022 16:39:28 GMT137ad170-7792-4731-bb38-c0d22fbe4515:6dc4f95d-8e0a-4c0a-b187-e40ea385e4b1Einar Thorsrud<p>Hi,</p> <p>The NVS API is synchronous, so when the call to nvs_write() returns, the data should have been written. I have confirmed with one of the developers that worked on the flash driver that this is the case, and he did not see any way the&nbsp;behavior&nbsp;you describe could happen. So unless there is an unknown bug, this should not be the case.</p> <p>Are you sure the DUT processed the data before it was powered down? Could there be an issue in your production script, for instance? (In any case, a solution could be to read back and verify that the data is properly written that way before powering down).</p><div style="clear:both;"></div>