This post, copied below, made in opensuse.org.help.install-boot-login forum got totally
corrupted via nntp. The code sections are missing, and the text is different than on the web.
On 2012-08-02 01:36, wichne wrote:
>
> Hello,
>
> I have had a similar problem. I am upgrading to openSUSE 12.1 on a
> system with a RAID 1 configuration and got:
> kernel (hd0)/vmlinuz-3.1.10-1.16-desktop root=/dev/md2 resume=/dev/md1
> splash=silent vga=0x31a
>
> Error 17: Cannot mount selected partition
>
> I could also not boot in failsafe.
>
> Here is the fdisk -l output.
>
> Since this is a RAID 1 setup, I activated the RAID:
> Two devices did not assemble: sda6 failed to be added to md4 because of
> an invalid argument; md4 thus tried to assemble from 0 drives and
> failed.
> sda5 failed to be added to md3, also because of an invalid argument;
> md3 thus tried to assemble from 0 drives and failed.
>
> I ran fdisk again:
>
> So now I see my partitions and I can identify them by contents as:
> md118 /
> md119 /data
> md120 /boot
> md121 /srv
> md125 /data (but files not touched since 2010…hmm…could my RAID
> have broken?)
>
> There is also supposed to be a /home and a /tmp partition, but those
> did not appear. Perhaps they are md3 and md4?
>
> So how do I change my bootloader command to point to the right
> partition? Switching it to hd1 didn’t work (not that I thought it
> would). How do I specify the RAID device?
>
> thanks,
> Bill
>
>
–
Cheers / Saludos,
Carlos E. R.
(from 11.4 x86_64 “Celadon” at Telcontar)
On Thu, 02 Aug 2012 01:02:00 +0000, Carlos E. R. wrote:
> What happened?
No idea, but I note that there are content differences in this post and
post #10 in the thread itself.
It’s possible that what happened was the poster edited the post and it
went through the gateway after the initial post but while it was being
edited. It may be that a post being edited across the gateway window
could be posted but since the gateway processed it, it wouldn’t be
processed on the next run through.
I’ll see if I have time over the weekend to test that theory out in my
sandbox.
On 2012-08-02 06:45, Jim Henderson wrote:
> On Thu, 02 Aug 2012 01:02:00 +0000, Carlos E. R. wrote:
>
>> What happened?
>
> No idea, but I note that there are content differences in this post and
> post #10 in the thread itself.
That’s the problem, yes.
> It’s possible that what happened was the poster edited the post and it
> went through the gateway after the initial post but while it was being
> edited. It may be that a post being edited across the gateway window
> could be posted but since the gateway processed it, it wouldn’t be
> processed on the next run through.
I thought of that, too.
> I’ll see if I have time over the weekend to test that theory out in my
> sandbox.
If that proves right, the edit window should close at the minute the gateway process runs (1
minute or 10 minutes, variable), or better, do not process the email till after 10 minutes have
really passed.
–
Cheers / Saludos,
Carlos E. R.
(from 11.4 x86_64 “Celadon” at Telcontar)
On Thu, 02 Aug 2012 08:33:21 +0000, Carlos E. R. wrote:
>> I’ll see if I have time over the weekend to test that theory out in my
>> sandbox.
>
> If that proves right, the edit window should close at the minute the
> gateway process runs (1 minute or 10 minutes, variable), or better, do
> not process the email till after 10 minutes have really passed.
That’s a far more difficult problem to solve, because the gateway runs on
a cron job, not as an event-driven process (ie, it doesn’t run 10 minutes
after someone makes a post, it runs at a regularly scheduled interval).
It doesn’t happen very often, and if I get a chance to completely rewrite
the gateway, my idea for doing so should resolve the issue by making it
event-driven rather than timed. But that’ll be a major rewrite, so it
won’t happen anytime soon.
On 2012-08-02 18:08, Jim Henderson wrote:
> On Thu, 02 Aug 2012 08:33:21 +0000, Carlos E. R. wrote:
>
>>> I’ll see if I have time over the weekend to test that theory out in my
>>> sandbox.
>>
>> If that proves right, the edit window should close at the minute the
>> gateway process runs (1 minute or 10 minutes, variable), or better, do
>> not process the email till after 10 minutes have really passed.
>
> That’s a far more difficult problem to solve, because the gateway runs on
> a cron job, not as an event-driven process (ie, it doesn’t run 10 minutes
> after someone makes a post, it runs at a regularly scheduled interval).
I know. So mark the post as non-editable on that cron job, even if that means 1 minute window
or no window. Which is pretty bad solution.
> It doesn’t happen very often, and if I get a chance to completely rewrite
> the gateway, my idea for doing so should resolve the issue by making it
> event-driven rather than timed. But that’ll be a major rewrite, so it
> won’t happen anytime soon.
Cest la vie… or something like that.
–
Cheers / Saludos,
Carlos E. R.
(from 11.4 x86_64 “Celadon” at Telcontar)
On Thu, 02 Aug 2012 19:28:06 +0000, Carlos E. R. wrote:
>> That’s a far more difficult problem to solve, because the gateway runs
>> on a cron job, not as an event-driven process (ie, it doesn’t run 10
>> minutes after someone makes a post, it runs at a regularly scheduled
>> interval).
>
> I know. So mark the post as non-editable on that cron job, even if that
> means 1 minute window or no window. Which is pretty bad solution.
It’s not that simple (it never is). The job just runs a php script that
takes messages from one side and posts them to the other and vice-versa.
And modifying the behaviour of vBulletin itself gets into supportability
issues (and an inability to apply security updates in a timely fashion).
IOW, it’s a non-trivial problem to solve with things as they currently
are.
>> It doesn’t happen very often, and if I get a chance to completely
>> rewrite the gateway, my idea for doing so should resolve the issue by
>> making it event-driven rather than timed. But that’ll be a major
>> rewrite, so it won’t happen anytime soon.
>
> Cest la vie… or something like that.
Indeed. When it comes to maintaining something like this, something that
affects 1 message in 10,000 (or more) isn’t likely to be a high priority
in anyone’s book. We’ve got far more pressing matters to address.
On 2012-08-02 23:00, Jim Henderson wrote:
> Indeed. When it comes to maintaining something like this, something that
> affects 1 message in 10,000 (or more) isn’t likely to be a high priority
> in anyone’s book. We’ve got far more pressing matters to address.
We don’t really know how many posts are affected. This time I noticed because the nntp post
made references to things that were not there and which I wanted to see. Other posts with
subtle edits, but important for the meaning, can go unnoticed.
Hopefully few.
–
Cheers / Saludos,
Carlos E. R.
(from 11.4 x86_64 “Celadon” at Telcontar)
On Thu, 02 Aug 2012 23:58:06 +0000, Carlos E. R. wrote:
> On 2012-08-02 23:00, Jim Henderson wrote:
>> Indeed. When it comes to maintaining something like this, something
>> that affects 1 message in 10,000 (or more) isn’t likely to be a high
>> priority in anyone’s book. We’ve got far more pressing matters to
>> address.
>
> We don’t really know how many posts are affected. This time I noticed
> because the nntp post made references to things that were not there and
> which I wanted to see. Other posts with subtle edits, but important for
> the meaning, can go unnoticed.
>
> Hopefully few.
It strikes me that if the meaning of the posts isn’t significantly
changed, then it’s insignificant. The fact that since these forums were
opened that this is the first (or possibly second) time that the question
has come up, it seems like a very, very minor thing.
On 2012-08-03 02:24, Jim Henderson wrote:
> On Thu, 02 Aug 2012 23:58:06 +0000, Carlos E. R. wrote:
>
>> On 2012-08-02 23:00, Jim Henderson wrote:
>>> Indeed. When it comes to maintaining something like this, something
>>> that affects 1 message in 10,000 (or more) isn’t likely to be a high
>>> priority in anyone’s book. We’ve got far more pressing matters to
>>> address.
>>
>> We don’t really know how many posts are affected. This time I noticed
>> because the nntp post made references to things that were not there and
>> which I wanted to see. Other posts with subtle edits, but important for
>> the meaning, can go unnoticed.
>>
>> Hopefully few.
>
> It strikes me that if the meaning of the posts isn’t significantly
> changed, then it’s insignificant. The fact that since these forums were
> opened that this is the first (or possibly second) time that the question
> has come up, it seems like a very, very minor thing.
The changed part can be sufficiently minor to avoid detection unless you compare both versions,
yet the nntp version can have quite a different meaning than what the web readers see.
I could do a demonstration right now.
Suppose I post, on minute 5, “I’m a bastard”. On minute 7 I edit it to say “I’m a saint”. Both
versions are independently congruent and undetectable - till you compare both, and yet they
have very different meaning.
–
Cheers / Saludos,
Carlos E. R.
(from 11.4 x86_64 “Celadon” at Telcontar)
On Fri, 03 Aug 2012 00:48:06 +0000, Carlos E. R. wrote:
> The changed part can be sufficiently minor to avoid detection unless you
> compare both versions,
> yet the nntp version can have quite a different meaning than what the
> web readers see.
>
> I could do a demonstration right now.
Sure, one could “game” the system to prove a point. My point is that in
normal operations, it’s been a very, very minor thing that has happened
and been reported very rarely.
Applying the Pareto rule, there are far, far more important things to be
addressing.
On 2012-08-03 02:57, Jim Henderson wrote:
> On Fri, 03 Aug 2012 00:48:06 +0000, Carlos E. R. wrote:
>> I could do a demonstration right now.
>
> Sure, one could “game” the system to prove a point. My point is that in
> normal operations, it’s been a very, very minor thing that has happened
> and been reported very rarely.
My point is not that the system can be fooled, but that if this happens it is difficult to notice.
> Applying the Pareto rule, there are far, far more important things to be
> addressing.
I hope that you are right, but I fear that you are not.
–
Cheers / Saludos,
Carlos E. R.
(from 11.4 x86_64 “Celadon” at Telcontar)
On Fri, 03 Aug 2012 01:08:06 +0000, Carlos E. R. wrote:
> My point is not that the system can be fooled, but that if this happens
> it is difficult to notice.
Noted.
>> Applying the Pareto rule, there are far, far more important things to
>> be addressing.
>
> I hope that you are right, but I fear that you are not.
As I said, it’s only been reported as an issue once or twice.
It’s a question of priorities, and as is stated in the FAQ, the NNTP
interface is a secondary interface.
The NNTP gateway was a custom third-party plugin created for an earlier
version of vBulletin that has been maintained by a couple of us to keep
this interface active on the Novell/SUSE/openSUSE forums sites. It’s not
like we can go to vBulletin and say “this needs to be fixed”, because
it’s not their code.
What I know of both how vBulletin and the gateway code work is that it’s
not a trivial issue to fix definitively without either changes to the
underlying vBulletin code (which would be unsupported/unsupportable, and
thus would never be implemented) or without a complete rewrite of the
gateway to properly deal with edits properly.
I have some ideas about how the gateway can be rewritten to accommodate
more event-driven processing, but it’s a matter of needing more hours in
the day to get everything done that I’m working on (including things I’m
actually paid to do, which have the priority). I have a list of
enhancements/fixes for minor issues for the gateway that I’ve added this
to.
I appreciate the report of the issue - but I don’t see a need for this to
turn into yet another endless debate.
As the readers of this sub-forum may have deduced, this thread has only a tangential relation to the issue of bootloaders and/or root partitions (a fair observation, as this sub-forum is regarding “Forums”), perhaps a title/subject change may be in order ?
Some are placed in the un-enviable position of reconciling the diverse connection needs of a common community, and are deserving of our support and trust.
On Fri, 03 Aug 2012 03:56:03 +0000, SeanMc98 wrote:
> As the readers of this sub-forum may have deduced, this thread has only
> a tangential relation to the issue of bootloaders and/or root partitions
> (a fair observation, as this sub-forum is regarding “Forums”), perhaps a
> title/subject change may be in order ?
Perhaps, though I don’t know that it’s overly necessary since there’s no
need for more discussion on the topic.
On 2012-08-03 04:03, Jim Henderson wrote:
> On Fri, 03 Aug 2012 01:08:06 +0000, Carlos E. R. wrote:
>
>> My point is not that the system can be fooled, but that if this happens
>> it is difficult to notice.
>
> Noted.
>
>>> Applying the Pareto rule, there are far, far more important things to
>>> be addressing.
>>
>> I hope that you are right, but I fear that you are not.
>
> As I said, it’s only been reported as an issue once or twice.
>
> It’s a question of priorities, and as is stated in the FAQ, the NNTP
> interface is a secondary interface.
…
I accept that you do not have resources to improve the situation. I accept that nothing can be
done at present. I’m not trying to push the issue.
But please understand that we really have no proof of how many posts are affected by it. Proof,
not feelings.
I’ll wait patiently till/if the gateway can be improved. I’m very glad that the gateway exists.
–
Cheers / Saludos,
Carlos E. R.
(from 11.4 x86_64 “Celadon” at Telcontar)
On 2012-08-17 18:29, Carlos E. R. wrote:
> On 2012-08-03 13:58, Carlos E. R. wrote:
>> On 2012-08-03 04:03, Jim Henderson wrote:
>
>>> As I said, it’s only been reported as an issue once or twice.
>
> I’ll try to record instances of the issue that I find. For example:
>
>
That post is different on each side. Via nntp it doesn’t make sense as the link is missing.
Unless it is later marked as spam, but in that case it is Linux oriented spam, at least, and
related to the thread in question (LVM recovery)
–
Cheers / Saludos,
Carlos E. R.
(from 12.1 x86_64 “Asparagus” at Telcontar)
