<div dir="ltr"><div class="gmail_default" style="font-size:small">Hi,<br></div><div class="gmail_extra"><br><div class="gmail_quote">On Fri, Jul 27, 2018 at 1:36 AM, Lars Ellenberg <span dir="ltr"><<a href="mailto:lars.ellenberg@linbit.com" target="_blank">lars.ellenberg@linbit.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span class="">On Mon, Jul 23, 2018 at 02:46:25PM +0200, Michal Michaláč wrote:<br>
> Hello,<br>
> <br>
> <br>
> <br>
> after replacing backing device of DRBD, content of DRBD volume (not only<br>
> backing disk) is invalid on node with inconsistent backing device, until<br>
> sync finishes. I think, correct behaviour is to read data from peer's<br>
> (consistent) backing device if process running on node with inconsistent<br>
> backing device wants to read unsynchronized part of DRBD volume.<br>
<br>
</span>...<br>
<br>
> If I skip create-md (step 4), situation is even worse - after attach disk,<br>
<span class="">> DRBD says volume is sychronized(!):<br>
> <br>
> log: Began resync as SyncTarget (will sync 0 KB [0 bits set])<br>
> <br>
> but after verification (drbdadm verify test), there are many out-of-sync<br>
> sectors. <br>
> <br>
> After disconnect/connect volume test, resync not started(!):<br>
> <br>
> log: No resync, but 3840 bits in bitmap!<br>
> <br>
> If I (on new DRBD volume) just disconnect -> write changes to primary -><br>
> connect, sync works correctly.<br>
<br>
> Versions (on both nodes are identical): <br>
> # cat /proc/drbd<br>
> version: 9.0.14-1 (api:2/proto:86-113)<br>
> GIT-hash: 62f906cf44ef02a30ce0c148fec223<wbr>b40c51c533 build by root@n2,<br>
> 2018-07-12 13:18:02<br>
> <br>
> Transports (api:16): tcp (9.0.14-1)<br>
> <br>
> # uname -a<br>
> Linux n2 4.15.17-1-pve #1 SMP PVE 4.15.17-9 (Wed, 9 May 2018 13:31:43 +0200)<br>
> x86_64 GNU/Linux<br>
> <br>
> # lvm version<br>
> LVM version: 2.02.168(2) (2016-11-30)<br>
> Library version: 1.02.137 (2016-11-30)<br>
> Driver version: 4.37.0<br>
<br>
> Is it bug or am I doing something wrong?<br>
<br>
</span>Thanks for the detailed and useful report,<br>
definetely a serious and embarassing bug,<br>
now already fixed internally.<br>
Fix will go into 9.0.15 final.<br>
We are in the progress of making sure<br>
we have covered all variants and lose ends of this.<br></blockquote><div><br></div><div style="font-size:small" class="gmail_default">Is this going to get back ported to 8.4 as well?</div><div style="font-size:small" class="gmail_default"><br></div></div></div></div>