Re: [mirrorbrain] wrong number of index expressions ?

From: Peter Pöml <peter_at_poeml.de>
Date: Fri, 31 Jan 2014 00:38:08 +0100
On Thu, Jan 30, 2014 at 12:47:11AM -0600, Robby Workman wrote:
> On Wed, 29 Jan 2014 22:53:51 +0100
> Peter Pöml <peter_at_poeml.de> wrote:
> 
> > Hi Robby,
> > 
> > sorry, a very late (probably too late) answer:
> > 
> > On Thu, Oct 10, 2013 at 11:22:26AM -0500, Robby Workman wrote:
> > > Okay, I'm probably missing something obvious, but any ideas on this?
> > 
> > I don't think you are missing something obvious. On the contrary, I
> > can't think of a rational reason why this error should occur. It's
> > definitely not a typical problem you run into...
> > 
> > One possibly related issue I find in Google is this one:
> > https://bugzilla.redhat.com/show_bug.cgi?id=993532
> > which is miscompilation of PostgreSQL. There is also a similar report
> > in a Gentoo forum. 
> > 
> > See comment #8 from Tom Lane:
> > https://bugzilla.redhat.com/show_bug.cgi?id=993532#c8
> > 
> > Other than that, no idea... But I suppose it's the above issue.
> > 
> > Did you find a solution?
> 
> 
> Well, maybe. :-)  It's all working fine now, but I don't know for 
> sure why or what fixed it.  I had a catastrophic failure of two 
> drives in the RAID array on that system, and while I was able to
> pull all of the data off of them, I had to do a completely new
> installation of the system.  The old version of pgsql was 9.1.3
> or some such (not sure exactly, but one of the early 9.1.x releases),
> and I installed 9.1.7 or .8, whichever was current at the time of
> the reinstall (I stayed with 9.1.x because I needed to get the 
> system back up ASAP and didn't want to mess with upgrading the db
> and introduce yet another potential point of failure).  Along the
> way, I used newer versions of all the mirrorbrain deps, to include
> the perl stuff, which involved much newer libwww-perl, which added
> lots of new deps that were no longer bundled with libwww-perl, and
> so on.  In a nutshell, this is why I wasn't sure where the problem
> actually was - it could be in any of those bits and pieces.  
> 
> I ended up downgrading everything except pgsql back to the older 
> versions, but it still didn't work.  This was about the time I
> mailed the list.  Within a day or so (maybe longer - I don't 
> recall exact timeline now), I noticed that there was a newer 
> pgsql release out (9.1.10), so I upgraded to it, and after
> restarting it and httpd, mirrorbrain was working.  I've since
> upgraded mirrorbrain to 2.17.0 as well as some of the deps, but
> I'm not touching the perl stuff again until I don't have a choice :-)
> 
> Anyway, I guess *maybe* the version of pgsql that I had first
> installed on the new system was problematic in some way (the
> system compiler is indeed gcc-4.8.2) and that 9.1.10 (or maybe .9)
> fixed that.  I'm just happy that it works :-)
> 
> -RW

Good, that all sounds like the same issue that I found here and there
popping up in Google, regarding miscompilation of PostgreSQL. The
dysfunctional PostgreSQL shipped with several distros and all seem to
have fixed the issue with an updated package later. So it wasn't a
problem in MirrorBrain. And it's fixed. Fine :-)

Thank you for your feedback!
Peter



_______________________________________________
mirrorbrain mailing list
Archive: http://mirrorbrain.org/archive/mirrorbrain/

Note: To remove yourself from this mailing list, send a mail with the content
 	unsubscribe
to the address mirrorbrain-request_at_mirrorbrain.org
Received on Thu Jan 30 2014 - 23:38:24 GMT

This archive was generated by hypermail 2.3.0 : Fri Jan 31 2014 - 01:17:04 GMT