Closed txf- closed 4 years ago
I remember testing on SPIFFS, but don't remember what the outcome was. I will try it out again and get back to you.
I tried the example on SPIFFS, but I did not face the problems you have mentioned. I could not complete the test because of hardware failure.
About heap values, my opinion is that they plateau once there are no more new sized data need to be allocated. They seem not to change because the same sized data gets allocated and deallocated.
See new library https://github.com/siara-cc/sqlite_micro_logger_arduino. It overcomes all memory problems and corruption issues. However there are limitations. For documentation see: https://github.com/siara-cc/sqlite_micro_logger_c.
Currently I'm trying to insert 1000+ entries into a db with a modified sqlite_bulk_data_insert.ino (to use SPIFFS).
I'm getting periodic disk I/O errors, seemingly at random, but with increasing frequency the larger the database becomes. Notably executing further inserts after the error presents no issues (until a certain limit on partition usage).
Looking at the heap values, they remain unchanged, could it be that the heap is getting fragmented?
Are periodic I/O errors expected?