Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
HI all, i resume this old post that didn't have an answer. I am experiencing the same identical problem. the 2 nodes are 2 KVM virtual machines on 2.6.30.5 that are supposed to work in dual primary mode with OCFS on top pf DRBD. The testbed is: - machines are up and running, synced and in dual primary mode. - when i shutdown -h one of the two, everything comes up correctly. - when i reboot (KVM is quite fast), i see that the rebooted node doesn't sync and disconnects. This is strange because i don't see any configuration problem. If i restart drbd, it comes up cleanly as well. How to fix and what's the problem ? This is the relevant portion of the log. [ 7.444342] drbd: initialised. Version: 8.3.2 (api:88/proto:86-90) [ 7.466196] drbd: GIT-hash: dd7985327f146f33b86d4bff5ca8c94234ce840e build by phil at fat-tyre, 2009-07-03 15:35:39 [ 7.489085] drbd: registered as block device major 147 [ 7.500701] drbd: minor_table @ 0xffff8800ddd5ac80 [ 7.523703] block drbd1: Starting worker thread (from cqueue [1691]) [ 7.535696] block drbd1: disk( Diskless -> Attaching ) [ 7.551927] block drbd1: Found 6 transactions (6 active extents) in activity log. [ 7.574272] block drbd1: Method to ensure write ordering: barrier [ 7.585844] block drbd1: max_segment_size ( = BIO size ) = 32768 [ 7.597424] block drbd1: drbd_bm_resize called with capacity == 104854328 [ 7.609724] block drbd1: resync bitmap: bits=13106791 words=204794 [ 7.632516] block drbd1: size = 50 GB (52427164 KB) [ 7.647568] block drbd1: recounting of set bits took additional 0 jiffies [ 7.659347] block drbd1: 0 KB (0 bits) marked out-of-sync by on disk bit-map. [ 7.671181] block drbd1: disk( Attaching -> UpToDate ) [ 7.682717] block drbd1: Barriers not supported on meta data device - disabling [ 7.773093] block drbd1: conn( StandAlone -> Unconnected ) [ 7.785603] block drbd1: Starting receiver thread (from drbd1_worker [1698]) [ 7.808437] block drbd1: receiver (re)started [ 7.824425] block drbd1: conn( Unconnected -> WFConnection ) [ 7.836513] block drbd1: bind before connect failed, err = -99 [ 7.852925] block drbd1: conn( WFConnection -> Disconnecting ) [ 7.868108] block drbd1: role( Secondary -> Primary ) [ 7.880114] block drbd1: Creating new current UUID [ 8.064182] block drbd1: Discarding network configuration. [ 8.075792] block drbd1: Connection closed [ 8.087158] block drbd1: conn( Disconnecting -> StandAlone ) [ 8.098898] block drbd1: receiver terminated [ 8.110451] block drbd1: Terminating receiver thread Thanks, Max Richard Hector ha scritto: > Hi all, > > One of my 2 machines doesn't seem to connect at boot time - doesn't > matter whether it's configured to come up as primary or secondary. This, > at a guess, seems relevant: > > ... > [ 29.196417] drbd0: conn( Unconnected -> WFConnection ) > [ 29.196417] drbd0: bind before connect failed, err = -99 > [ 29.196417] drbd0: conn( WFConnection -> Disconnecting ) > ... > > I've tried to find any docs regarding this without success - that bind > is an internal kernel one, not bind(2) (the syscall), right? My > expertise at navigating the kernel source is rather limited ... > > If I run /etc/init.d/drbd restart after boot, it comes up fine. > > Both machines are Debian Lenny, amd64, with drbd 8.3 from backports.org. > > Any suggestions? Any more useful info I can supply? >