On Wednesday, April 18, 2018 at 2:49:37 PM UTC-5, jmp242 wrote:
>
> I just got back to this. I tried re-installing the Windows 10 VM, tried a
> different hostgroup in Foreman, seems like it's a catalog problem of some
> sort.
> puppet agent --trace gives no output for me.
>
> puppet agent --debug around the error gives:
> Debug: Prefetching chocolatey resources for package
> Debug: Executing: 'C:\ProgramData\chocolatey\bin\choco.exe pin list'
> Debug: Executing 'C:\ProgramData\chocolatey\bin\choco.exe list -lo -r'
> Debug: Executing 'C:\ProgramData\chocolatey\bin\choco.exe upgrade --noop
> chocolatey.extension -r'
> Debug: Class[Yum]: The container Stage[main] will propagate my refresh
> event
> Debug: Prefetching windows resources for package
> Debug: Unable to convert value
> "{\x00B\x003\x00A\x00F\x00C\x006\x000\x008\x00-\x00D\x008\x001\x001\x00-\x000\x000\x000\x003\x00-\x004\x000\x005\x000\x00-\x002\x001\x00F\x00B\x002\x005\x00B\x004\x008\x00D\x006\x00E\x00}\x00\x00\x00\x00\x00{\xEEs\vj\x93\x00\x00H\x14t\x00\x18.t\x00d\x00l\x00l\x00\x00\x00~\xEEr\x0Fb\x93\x00\b\xD0\xC3\xF2w`8t\x00\x00
>
> \xD9u\b
> \xD9ul~\xD8uX\xDC\x8Bu\x00\x00\x00\x00\x00\x000\x000$t\x00\x02\x00\x00\x00v\xEEr\ag\x93\x00\f?\x0E\x00\x00\t\x04\x00\x00\x00\x00\x00\x00\x01\x00\x00\x00\x00\x00\x00\x00(:t\x00\xB0:t\x00@\xCDs\x00
>
> \xCCs\x00(Kt\x00hMt\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00(;t\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\xE8\x03\x00\x00\x00\x00\x00\x00\xD8Dt\x00\x06\x03\x00\x00"
> .force_encoding("UTF-16LE") to encoding UTF-8 due to
> #<Encoding::InvalidByteSequenceError:
> "X\xDC" on UTF-16LE>
> Debug: Storing state
> Debug: Stored state in 0.08 seconds
> Error: Failed to apply catalog: "X\xDC" on UTF-16LE
>
>
> Notepad++ shows the manifests used here (as far as I'm able to tell) are
> both UTF-8...
>
When I said the problem is more likely to be in the manifests than in
Puppet itself, I didn't mean specifically that the manifests might be
encoded wrongly, although there are potential issues of that kind. That
Notepad++ thinks they are encoded in UTF-8 is a relevant detail, but not
indicative of anything one way or the other.
The debug messages you presented are more useful, though. Supposing that
you haven't cut anything out, it looks like you have a chunk of *binary*
data that Puppet is for some reason trying to interpret as text. The
embedded sequences of multiple consecutive 0 bytes are a dead giveaway that
it's binary. I'm uncertain whether Puppet is expecting from context that
the byte sequence is UTF-16-encoded text, or whether it is guessing based
on the content, but the byte sequence overall does not conform to any of
the UTFs, and it does not look like text in any of ASCII-compatible
single-byte encoding, either.
Since this appears to be happening while Puppet is pre-fetching resources,
I am forced to revise my opinion that the problem is somewhere among your
manifests. Your manifests are probably not responsible for this, except
inasmuch as they declare at least one Package resource that explicitly or
implicitly uses the default 'windows' provider. That's a flaw only if you
intended to do differently.
Inasmuch as the problem appears to occur in the scope of prefetching
'windows' packages, it is probably right to attribute it to Puppet. But
Puppet's 'windows' package provider generally works as advertised, so it
seems likely that something about the current state of the target machine
is contributing.
John
--
You received this message because you are subscribed to the Google Groups
"Puppet Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to puppet-users+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/puppet-users/fa856915-87a7-4bf8-8173-2474e6cc9f7a%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
