Open luckyho opened 4 weeks ago
Hi, I suggest using the following for debugging purposes:
rbd-wnbd map rendersphere/test1-img01 -n client.rendersphere -f
-f
tells it to run as a daemon and map the device in the same process instead of using the centralized service, thus it's easier to obtain the actual error.
ok, i will try it
when adding -f to run , it works well. but If -f is not added, especially after the system is restarted, the probability of map failure will be very high
There are some logs below
it looks like timeout?
the rbd server is start at 12:37
Interesting, it does look like a timeout. This old version of Ceph (Pacific) spawns a separate process for each mapped image and for some reason it fails in this case.
Newer versions use a single daemon process per host and include many other bug fixes that have not been ported to Pacific, which is no longer maintained. Would it be possible to upgrade to Ceph Squid or Reef?
Mount rbd on win10, rbd wnbd returns -22 The screenshot is below
ceph version : 15.2.17 rbd wnbd version : 16.4.10