[DRBD-user] RE: How to do partial re-sync -> Sorry for not being detailed enough

Lars Ellenberg Lars.Ellenberg at linbit.com
Thu Aug 18 20:50:07 CEST 2005

Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.


/ 2005-08-18 21:23:38 +0800
\ Vic Berdin:
> Hi Lars,
> 
> > ------------------------------
> > I am sure drbd is fine with any library and bash compatible shell.
> > I rather suspect you doing something "stupid" (sorry for directness).
> 
> 1. First of, my environment does not have "/bin/bash" (and I do not plan to
> use "bash"). Instead, all it has is /bin/sh -> /bin/busybox. Thus, I changed
> all calls to /bin/bash.
> 
> 2. My ip -> /bin/busybox doesn't support route_module, thus, it spurrs "No
> route from me (192.168.2.2) to peer (192.168.2.1) messages as I run my
> /etc/rc.d/init.d/drbd start (from scripts/drbd; also tweaked to work on my
> environment). So I made some tweaks to prioritize ifconfig -> /bin/busybox
> and augment the absence of `ip route`.
> 
> Note: Although the drbd.o module loads and /proc/drbd is available even if I
> do not do (2). I just thought I had to do it just to be sure/and be free of
> these "No route..." errors.
> 
> > if you talk "bash" to me, that is,
> > starting from no drbd module loaded,
> > step by step the commands that you issue,
> > indicating on which node, and your expectations,
> > including /proc/drbd and/or dmesg | tail | grep -i drbd
> > (if it changed from the step before),
> > then probably the list or I can point out where the actual thinko is.

...

> Now, after I unmount the target device on Secondary, and re-enable DRBD on
> it, once more, as Secondary, while the other node remains Primary, I did
> some dumping of data on Primary's source device (simple `cp`+`sync` commands
> just to add data on it and made sure they're not just in buffer).
> 

what is "Primary's source device" ?
can you please show the mounts?
and the actual cp and sync commands entered?

> And while I'm at it, I'm also doing a series of "cat /proc/drbd
> >>/tmp/copying.res;" and "dmesg|tail|grep -i drbd>>/tmp/copying.res" on
> Secondary. After HDD LED on Primary stops blinking, I openned the file, and
> here're the results:
> 
> version: 0.7.11 (api:77/proto:74)
> SVN Revision: 1807 build by root at linuxmachine, 2005-08-12 10:20:57
>  0: cs:Connected st:Secondary/Primary ld:Consistent
>     ns:0 nr:0 dw:0 dr:0 al:0 bm:0 lo:0 pe:0 ua:0 ap:0

you are not writing to the drbd device.
I have no Idea where you are writing to,
but you do not use the drbd device.

> Note that the results are the same all through out the "copying_on_primary"
> event/process until it finishes. If I am already able to perform full-sync
> each time I want to, what stops DRBD from performing partial-sync during a
> "Connected" state?

you are not using it.

> > or, to start with,
> > just point out were the "drbd quickstart guide" does not work for you.
> 
> Your guides/how to are fine. DRBD is fine, I am sure. That is why I'm
> guessing/trying to figure out if my thin environment is the culprit for this
> failure.

thanks.
so does this mean you actually followed it through,
and it did work for you, or are you just assuming that it would work?

> And by the way, if it matters, I'm implementing DRBD on the Nodes' internal
> device. That is, I do not have the actual hostnames (`hostname`) as part of
> my /etc/hosts.

this paragraph I do not understand.

-- 
: Lars Ellenberg                                  Tel +43-1-8178292-0  :
: LINBIT Information Technologies GmbH            Fax +43-1-8178292-82 :
: Schoenbrunner Str. 244, A-1120 Vienna/Europe   http://www.linbit.com :
__
please use the "List-Reply" function of your email client.



More information about the drbd-user mailing list