Re: vacuumdb question/problem

  Tom,

It was not consuming any CPU time. I found that another process on the
machine failed. That process was trying to write to a different table
(not the one I was vacuuming) in the database and that table was locked
(not sure why).  It produced thousands of errors which caused the log
files to grow large enough that the entire disk was filled, and this
brought everything to a halt.

So I had to kill my process, recover disk space and get the machine back
in working condition for the weekend. I guess I will attempt to do the
full vacuum again next week.

I am still wondering how the vacuum process actually works. When it
throws the output lines that show how many rows are
recoverable/nonremovable, does this mean that the vacuum has completed?
Or do those output lines show the vacuum has scanned through the
database...found what what it is going to recover and then takes on the
vacuum process?

Thanks,
Dave

Tom Lane said the following on 7/15/2011 1:12 PM:
> Simon Riggs<simon@2ndQuadrant.com>  writes:
>> On Fri, Jul 15, 2011 at 5:10 PM, David Ondrejik<David.Ondrejik@noaa.gov>  wrote:
>>> Since then, the process has continued to run (for about 20 hrs) without any
>>> additional information being returned.
>> Probably locked behind another long running task that is holding a buffer pin.
> That's possible, or it could be busy vacuuming some (really large?)
> index.  Is the process actually busy, as in consuming CPU time according
> to top or other process monitoring tool?
>
>             regards, tom lane