Linus, Ingo:
the attached patch, modifies a warning message in md.c which seems to
often cause confusion - the following email includes one example
there-of (there have been others over the months).
Hopefully the new text is clearer.
(patch against 2.4.0-test11-pre7)
NeilBrown
On Sunday November 19, [email protected] wrote:
> On Sun, Nov 19, 2000 at 03:39:43AM -0800, George Garvey wrote:
> > Is this something to be concerned about? It sounds like a disaster waiting
> > to happen from the message. This is on 2 systems (with similar disk setups
> > [same other than size]).
>
> > Nov 18 16:31:02 mwg kernel: md: serializing resync, md1 has overlapping physical units with md2!
>
> Nope, nothing to worry about -- it's just a bad choice of wording ;)
>
> What it means is that some partititions in md1 and md2 are on the same disk,
> and that the md-code will not do the reconstruction of these arrays in
> parallel [of course, for performance reasons].
>
--- ./drivers/md/md.c 2000/11/20 00:33:08 1.2
+++ ./drivers/md/md.c 2000/11/20 00:44:19 1.3
@@ -3279,7 +3279,7 @@
if (mddev2 == mddev)
continue;
if (mddev2->curr_resync && match_mddev_units(mddev,mddev2)) {
- printk(KERN_INFO "md: serializing resync, md%d has overlapping physical units with md%d!\n", mdidx(mddev), mdidx(mddev2));
+ printk(KERN_INFO "md: serializing resync, md%d has shares one or more physical units with md%d!\n", mdidx(mddev), mdidx(mddev2));
serialize = 1;
break;
}
On Mon, 20 Nov 2000, Neil Brown wrote:
> the attached patch, modifies a warning message in md.c which seems to
> often cause confusion - the following email includes one example
> there-of (there have been others over the months).
>
> Hopefully the new text is clearer.
yep, agreed - thanks!
Ingo
Folks,
I won't try and invent a patch(1) for this, but might I suggest changing:
md: serializing resync, md%d has shares one or more physical units with
md%d!\n
to
md: serializing resync, md%d shares one or more disk drives with md%d.
Array performance may suffer.\n
Regards,
Ruth
At 12:49 AM 11/20/00, you wrote:
> the attached patch, modifies a warning message in md.c which seems to
> often cause confusion - the following email includes one example
> there-of (there have been others over the months).
> > What it means is that some partititions in md1 and md2 are on the same
> disk,
> > and that the md-code will not do the reconstruction of these arrays in
> > parallel [of course, for performance reasons].
> >
>
>
>--- ./drivers/md/md.c 2000/11/20 00:33:08 1.2
>+++ ./drivers/md/md.c 2000/11/20 00:44:19 1.3
>@@ -3279,7 +3279,7 @@
> if (mddev2 == mddev)
> continue;
> if (mddev2->curr_resync && match_mddev_units(mddev,mddev2)) {
>- printk(KERN_INFO "md: serializing resync, md%d has
>overlapping physical units with md%d!\n", mdidx(mddev), mdidx(mddev2));
>+ printk(KERN_INFO "md: serializing resync, md%d has
>shares one or more physical units with md%d!\n", mdidx(mddev), mdidx(mddev2));
> serialize = 1;
> break;
> }
>-
--
Ruth
Ivimey-Cook [email protected]
Technical
Author, ARM Ltd [email protected]