Does pkg automatically download INDEX?

[Naram Qashat]

On 08/15/14 13:14, Naram Qashat wrote:
> On 08/12/14 14:11, Patrick Powell wrote:
>> On 08/09/14 18:15, Naram Qashat wrote:
>>> On 08/09/14 19:45, Scot Hetzel wrote:
>>>> On Fri, Aug 8, 2014 at 7:00 PM, Naram Qashat <cyberbotx at cyberbotx.com> wrote:
>>>>> On 08/04/14 07:28, David Wolfskill wrote:
>>>>>>
>>>>>> On Mon, Aug 04, 2014 at 07:09:33AM -0400, Naram Qashat wrote:
>>>>>>>
>>>>>>> On 08/03/14 22:14, David Wolfskill wrote:
>>>>>>>>
>>>>>>>> On Sun, Aug 03, 2014 at 10:10:27PM -0400, Naram Qashat wrote:
>>>>>>>>>
>>>>>>>>> ...
>>>>>>>>> If there is
>>>>>>>>> a way to find out when any process is attempting to modify a file, that
>>>>>>>>> would
>>>>>>>>> probably help me narrow it down, but I'm not aware of anything that can
>>>>>>>>> do that,
>>>>>>>>> ...
>>>>>>>>
>>>>>>>>
>>>>>>>> Well, "chflags schg /usr/ports/INDEX*" would *prevent* the modification
>>>>>>>
>>>>>>> ...
>>>>>>> This was a really good suggestion.....
>>>>>>
>>>>>>
>>>>>> Glad to help.  :-)
>>>>>>
>>>>>> Peace,
>>>>>> david
>>>>>
>>>>>
>>>>> OK, so while no programs have whined or complained, I get the feeling that
>>>>> something on my system is running portsnap without my knowledge. When I had
>>>>> set the schg flag on INDEX-9, an INDEX-9.bz2 file came up. I set the schg
>>>>> flag on that as well, and now I notice there are a bunch of files called
>>>>> .fetch.??????.INDEX-9.bz2 (where ?????? is a random string), as well as a
>>>>> file called .portsnap.INDEX. As far as I know, I don't have anything
>>>>> configured to run portsnap, but is there something that defaults to running
>>>>> portsnap occasionally? I couldn't find anything that would do that.
>>>>>
>>>>
>>>> Do your have a crontab entry that is running portsnap with the -I
>>>> (update INDEX) option?
>>>>
>>>> http://www.pl.freebsd.org/doc/handbook/portsnap.html
>>>
>>> As far as I can tell, no, none of my crontabs have any references to portsnap
>>> in them. This is making me a bit stumped as to why it would be happening. I
>>> checked the main /etc/crontab, I checked the crontabs in /var/cron/tabs. I
>>> have searched inside of /etc and /usr/local/etc for anything related to
>>> portsnap. Nothing that would be doing this is coming up at all.
>>>
>>> Thanks,
>>> Naram Qashat
>>> _______________________________________________
>>> freebsd-ports at freebsd.org mailing list
>>> http://lists.freebsd.org/mailman/listinfo/freebsd-ports
>>> To unsubscribe, send any mail to "freebsd-ports-unsubscribe at freebsd.org"
>>>
>> I ran into something similar once,  and found out what was happening this way.
>>
>> 1.  replace the portsnap executable with a shell script.  Rename portsnap to
>> something
>>       like /usr/sbin/portsnap.orig
>> 2.  This shell script should dump the current ENV and other stuff to a log file.
>> Don't forget to put in a timestamp.
>> And then do:
>>    exec /usr/sbin/portsnap.orig $*
>>
>> I did this and found that there was something in one of the .login scripts.
>> Grrrr...
>
> I really liked this suggestion, and did just that. Unfortunately, it seems that
> portsnap may not be the culprit here, as I haven't had any log files created
> from my modified script nor is there a .portsnap.INDEX file being created after
> I deleted the one that was there, but I still have a bunch of
> .fetch.??????.INDEX-9.bz2 files in /usr/ports. I've been trying to search for
> anything on my system that even references INDEX-9, but I can't find anything
> else that would cause this to happen.
>
> Thanks,
> Naram Qashat

Well, I finally figured out what the culprit was. Webmin was configured to do 
automatic updates and it was running "make fetchindex" on the ports tree. I 
disabled that and now it has stopped trying to download the index. I only 
managed to track that down after noticing that the fetchindex target was the 
only other thing that could have downloaded the index file, and modifying 
/usr/ports/Makefile so it would write the contents of 
/proc/${.MAKE.PPID}/cmdline to a file so I could see what was invoking make. The 
.MAKE.PPID variable was pointed out to me on IRC.

Thanks,
Naram Qashat