On Sep 16, 2009, at 12:21 AM, R.I.Pienaar wrote:
>
> hello,
>
> ----- "Luke Kanies" <[email protected]> wrote:
>
>
>>
>> How does #2615 corrupt data?
>
> The data that 0.24.8 produce is not the same as 0.25.0 from a
> template, is this not corruption?
>
> # hexdump /var/lib/puppet/clientbucket/1/b/
> 6/3/8/0/1/4/1b6380149890b7ae59ebef53682abe93/contents|tail -n 2
> 0000870 6972 746e 4020 6164 6174 0a3b 0a7d 000a
> 000087f
>
> # hexdump /usr/local/bin/machdb-send|tail -n 2
> 0000870 6972 746e 4020 6164 6174 0a3b 0a7d 0a0a
> 0000880
>
>
>> This bug just causes services to get restarted when they shouldn't,
>> right? I know this is a problem, but you should only be doing this
>
> you must have missed my most recent updates to the bug.
Indeed I did; I've renamed the ticket again to make the problem more
clear, hopefully.
>
>> kind of upgrade during a maint. window or something anyway, and
>> services should either be able to restart cleanly or should not be
>> restarted by Puppet.
>
> I don't think we need to assume that a puppet upgrade on nodes has
> to be done in a maintenance window, one version of puppet should
> produce the same result as another version of puppet, so what's the
> risk in just updating puppet? If puppet 0.24.8 said all resources
> were in their expected state then so should 0.25.0 if the source
> material does not change so no maintenance window should be needed.
>
> Also assuming 0.24.8 didn't have a corruption bug of its own, but I
> don't think in this case that's what happened.
I'm not saying that 0.25 behaved correctly in this case, I'm just
saying I wouldn't tend to do this kind of upgrade unless I was either
very sure it was correct or I was doing it at a time when I could
afford a bit of strangeness, especially given that the release just
came out so recently.
>
>> I don't mean to downplay the importance of the bug, but we can't
>> reproduce it and you thought it was minimal enough that you've
>> continued deploying 0.25 on all of your machines even with this bug.
>
> Two things, I thought when I first opened the bug that it was
> minimal, it isn't, I was wrong, I wouldn't have rolled out 0.25.0 if
> I understood the real impact.
>
> The machines I updates are not a critical bunch of machines, so no
> biggie if it was just restarting services when it shouldn't, but
> since this bug is changing data I'll be spending today going through
> my old logs and finding every instance on every machine and
> verifying the appended data didn't cause a problem :(
Ok, I see.
--
If I want your opinion, I'll read your entrails.
--Doug Shewfelt
---------------------------------------------------------------------
Luke Kanies | http://reductivelabs.com | http://madstop.com
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups
"Puppet Developers" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to
[email protected]
For more options, visit this group at
http://groups.google.com/group/puppet-dev?hl=en
-~----------~----~----~----~------~----~------~--~---
