On Thu, 20 May 2010, Fred Ar wrote: > Hello, > > today I noticed some strange things on my 3 machines with each 1 hard > disk, 1 osd, 1 mds, 1 mon. They are named r1-9, r1-10, r1-11. > [...] > why do r1-9 and r1-10 appear twice, and why does r1-11 appear 5 times? Is it a problem, how to fix it ? > > I also had a 2 blacklisted OSDs (out of 3). Could it be because I > restarted them many times in a row? How to clear the blacklisted status > by hand (if I don't want to wait the delay)? Yeah. The monitor doesn't clear our 'laggy' (unresponsive, possibly dead) mds entries if they are standby, since they might come back and nothing is depending on them. We can add a longer timeout to clean these out if its clear the system is otherwise functional (e.g., other mds's are responsive). The blacklist entries can't be cleaned out manually right now, but there is no significant 'cost' to them being there, so you can ignore them. The blacklist just means that any IO they attempt with the OSDs will be rejected, to avoid a partitioned or misbehaving or laggy daemon from fighting with another mds instance. sage -- To unsubscribe from this list: send the line "unsubscribe ceph-devel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
