When I connect to a Network Block Device (NBD) using qemu-nbd -c /dev/nbd0 /tmp/disk and disconnect subsequently without any modifications via qemu-nbd -d /dev/nbd0, I get the following error in my syslog:
Jun 17 05:40:28 os-compute01 kernel: [49031.670805] nbd0: p1
Jun 17 05:40:28 os-compute01 kernel: [49031.917155] block nbd0: NBD_DISCONNECT
Jun 17 05:40:28 os-compute01 kernel: [49031.917303] block nbd0: Receive control failed (result -32)
Jun 17 05:40:28 os-compute01 kernel: [49031.917502] block nbd0: queue cleared
The two commands succeed and have exit code 0. I am running an Ubuntu Server 14.04.2 LTS on Kernel 3.16.0-41. I have also tried Kernel 3.13.0-55 resulting in the same output. Actually I am not sure, if I have a problem or if the output is just misleading.
The use case can be easily reproduced. Can someone help me to clarify, wheter I have a problem on my machine or not. My os-compute nodes are compute nodes for the current OpenStack Kilo release.
This is perfectly normal. When you disconnect, the kernel keeps reading until the connection is gone. That then produces that message. You can safely ignore it.