we are encountering a problem when running analytics on Toradex Apalis TK1 + IXORA V1.1 platform… after a while (about 2 minutes), we start getting kernel errors of the form: mmcblk0: error -110 sending status command at a rapid rate. From that point on the system is unustable and needs to be rebooted. the same scenario does not cause a problem on our reference platform - the NVIDIA Jetson TK1 dev kit. Both platforms have the exact same jetpack version. mmcblk0 is the main mmc storage (not the sd card) which includes the root partition and main storage (e.g. when we or the OS write something to a folder that’s not on the SD). I tried stress testing the system using another program and didn’t see this problem. Any suggestions?
What exact hardware (module and carrier board) and software versions are you talking about?
I guess it could be related to the power supply not providing enough current which in term could be a weak external power source, a carrier board with a too weak power supply resp. not enough bulk capacitance or the later on the module itself.