Closed gsleap closed 3 years ago
Further investigation shows that the issue is, the command to launch dada_diskdb, after hitting a timeout of 32 secs, the process does not die. This causes multiple dada_diskdb calls (all the same) to build up. Need to ensure the dada_diskdb process get's killed properly
Closed by ca4588270a2aed03e77826f745f8918e67e29716
Currently, if there is no free space in the ringbuffer, the same item will be added to the queue each time it tries. This results in lots of copies of the same item lining up to go in the ring buffer.
NOTE: order is critical so best to fail, but not requeue otherwise it might requeue to the end and then our observation is out of order.