[BALUG-Test] Oooh, ... archive tests ... :-)

Wed Sep 20 18:16:02 PDT 2017

> From: "Rick Moen" <rick@linuxmafia.com>
> Subject: Re: [BALUG-Test] Oooh, ... archive tests ... :-)
> Date: Wed, 20 Sep 2017 14:35:12 -0700

> Aaron, I seem to no longer have Michael's upthread posting around,
> so I'm keying off your post to respond to him.
>
> Michael Paoli wrote:
>
>> I changed the archiving
>> of BALUG-Test from (default of) monthly, to yearly,
>> that's all fine and good *except* ... archiving looks
>> a bit funky ... notably changing that setting doesn't
>> retroactively change what's already been archived, so,
>> e.g. on
>> https://temp.balug.org/pipermail/balug-test/
>> we see something that looks approximatley like:
>> Archive ...
>> August 2017: ...
>> July 2017: ...
>> 2017: ...
>
> Fix that using /var/lib/mailman/bin/arch to re-gen the archive, like:
>
> $ su -
> # su - list
> $ cd /var/lib/mailman
> $ bin/arch -q --wipe balug-test  
> archives/private/balug-test.mbox/balug-test.mbox
> $ exit
> # exit
> $
>
> You don't want to be the root user, or you'll end up having to later
> retroactively chown to list:list the generated archves.  (I've made that
> mistake a time or three.  Maybe you just figured that out.  ;->  )
>
> /var/lib/mailman/bin/arch will set permissions right.  Just don't run it
> as the wrong user.
>
> (The 'arch' in $PATH is of course something else entirely.)

Been there, done that.  ;-)
Though the bit about it being applied (done) to the other lists
was rather left as implied by that posting, but yes, also done.

references/excerpts:

https://temp.balug.org/pipermail/balug-test/2017/000014.html

> From: "Michael Paoli" <Michael.Paoli@cal.berkeley.edu>
> To: BALUG-Test <balug-test@temp.balug.org>
> Subject: Oooh, ... archive tests ... :-)
> Date: Mon, 18 Sep 2017 01:39:58 -0700

> So ... since some while (week(s)?) back, I changed the archiving
> of BALUG-Test from (default of) monthly, to yearly,
> that's all fine and good *except* ... archiving looks
> a bit funky ... notably changing that setting doesn't
> retroactively change what's already been archived, so,
> e.g. on
> https://temp.balug.org/pipermail/balug-test/
> we see something that looks approximatley like:
> Archive ...
> August 2017: ...
> July 2017: ...
> 2017: ...
> Which does look kind'a odd (after all, those months are also within
> the year 2017 - and it also may not be immedately visually apparent
> as to whether or not the 2017 archive does or doesn't include
> July 2017 and August 2017.
> Also, archive generally lists most recent at top, so,
> having the most recent (year) data at the bottom is also
> rather odd in its presentation.
>
> So ... let's reload those archives (I'm overdue for some testing
> of that anyway) ...
>
> $ hostname && id && ls -ld  
> /var/lib/mailman/archives/private/balug-test.mbox/balug-test.mbox  
> /var/lib/mailman/archives/private/balug-test.mbox/  
> /var/lib/mailman/archives/private/ && umask
> balug-sf-lug-v2.balug.org
> uid=38(list) gid=38(list) groups=38(list)
> drwxrws--- 12 www-data list  4096 Aug  6 03:11  
> /var/lib/mailman/archives/private/
> drwxrwsr-x  2 root     list  4096 Jul 10 01:13  
> /var/lib/mailman/archives/private/balug-test.mbox/
> -rw-rw-r--  1 list     list 63407 Aug 19 23:53  
> /var/lib/mailman/archives/private/balug-test.mbox/balug-test.mbox
> 0077
> $
> //I'd not altered permissions or used the mailman arch command to (re)load an
> //archive ... that's the "Debian way" they're set up - at least on
> //old-stable.  Note also that that differs somewhat from what the
> //mailman check_perms says they ought be ... but regardless it works,
> //and I'm guestimating for Debian is (moderately?) preferable settings
> //on those.
> $ (umask 0114 && /usr/lib/mailman/bin/arch --wipe balug-test)
> ...
> And ...
> https://temp.balug.org/pipermail/balug-test/
> that looks *much* better,
> ... but ... looks can be deceiving,
> web page looks fine, but web accessible archive bits thereof ...
> nope - looks like it only has what was already in the "year"
> archive, but lacks the earlier messages.
> Ah, ... silly browser (and/or what mailman may have earlier
> told it regarding caching) ... a reload of the page, and then
> all looks fine.
> And ... permissions?
> Let's see if what resulted looks good/sane in that regard,
> most notably expecting 'em to match similar archives - or at least
> be highly compatible (most notably relevant access via the group list).
>
> $ (cd /var/lib/mailman/archives/ && find * -follow -type f ! -mmin  
> +60 -exec ls -ld \{\} \;)
> -rw-rw-r-- 1 list list 1143 Sep 18 01:07 private/balug-test/index.html
> -rw-rw-r-- 1 list list 31152 Sep 18 01:07 private/balug-test/2017.txt
> -rw-rw-r-- 1 list list 5036 Sep 18 01:07 private/balug-test/2017/000003.html
> -rw-rw-r-- 1 list list 6457 Sep 18 01:07 private/balug-test/2017/000005.html
> -rw-rw-r-- 1 list list 2185 Sep 18 01:07 private/balug-test/2017/000000.html
> -rw-rw-r-- 1 list list 3804 Sep 18 01:07 private/balug-test/2017/000007.html
> lrwxrwxrwx 1 list list 11 Sep 18 01:07  
> private/balug-test/2017/index.html -> thread.html
> -rw-rw-r-- 1 list list 3449 Sep 18 01:07 private/balug-test/2017/date.html
> -rw-rw-r-- 1 list list 7136 Sep 18 01:07 private/balug-test/2017/000001.html
> -rw-rw-r-- 1 list list 2522 Sep 18 01:07 private/balug-test/2017/000008.html
> -rw-rw-r-- 1 list list 2740 Sep 18 01:07 private/balug-test/2017/000010.html
> -rw-rw-r-- 1 list list 8441 Sep 18 01:07 private/balug-test/2017/000002.html
> -rw-rw-r-- 1 list list 8244 Sep 18 01:07 private/balug-test/2017/000004.html
> -rw-rw-r-- 1 list list 2553 Sep 18 01:07 private/balug-test/2017/000009.html
> -rw-rw-r-- 1 list list 7865 Sep 18 01:07 private/balug-test/2017/000006.html
> -rw-rw-r-- 1 list list 3443 Sep 18 01:07 private/balug-test/2017/subject.html
> -rw-rw-r-- 1 list list 2416 Sep 18 01:07 private/balug-test/2017/000011.html
> -rw-rw-r-- 1 list list 3390 Sep 18 01:07 private/balug-test/2017/000012.html
> -rw-rw-r-- 1 list list 3994 Sep 18 01:07 private/balug-test/2017/thread.html
> -rw-rw-r-- 1 list list 3075 Sep 18 01:07 private/balug-test/2017/000013.html
> -rw-rw-r-- 1 list list 3445 Sep 18 01:07 private/balug-test/2017/author.html
> -rw-rw---- 1 list list 1942 Sep 18 01:07  
> private/balug-test/database/2017-thread
> -rw-rw---- 1 list list 19497 Sep 18 01:07  
> private/balug-test/database/2017-article
> -rw-rw---- 1 list list 2455 Sep 18 01:07  
> private/balug-test/database/2017-subject
> -rw-rw---- 1 list list 1756 Sep 18 01:07  
> private/balug-test/database/2017-date
> -rw-rw---- 1 list list 2014 Sep 18 01:07  
> private/balug-test/database/2017-author
> -rw-rw---- 1 list list 847 Sep 18 01:07 private/balug-test/pipermail.pck
> -rw-rw-r-- 1 list list 1143 Sep 18 01:07 public/balug-test/index.html
> -rw-rw-r-- 1 list list 31152 Sep 18 01:07 public/balug-test/2017.txt
> -rw-rw-r-- 1 list list 5036 Sep 18 01:07 public/balug-test/2017/000003.html
> -rw-rw-r-- 1 list list 6457 Sep 18 01:07 public/balug-test/2017/000005.html
> -rw-rw-r-- 1 list list 2185 Sep 18 01:07 public/balug-test/2017/000000.html
> -rw-rw-r-- 1 list list 3804 Sep 18 01:07 public/balug-test/2017/000007.html
> lrwxrwxrwx 1 list list 11 Sep 18 01:07  
> public/balug-test/2017/index.html -> thread.html
> -rw-rw-r-- 1 list list 3449 Sep 18 01:07 public/balug-test/2017/date.html
> -rw-rw-r-- 1 list list 7136 Sep 18 01:07 public/balug-test/2017/000001.html
> -rw-rw-r-- 1 list list 2522 Sep 18 01:07 public/balug-test/2017/000008.html
> -rw-rw-r-- 1 list list 2740 Sep 18 01:07 public/balug-test/2017/000010.html
> -rw-rw-r-- 1 list list 8441 Sep 18 01:07 public/balug-test/2017/000002.html
> -rw-rw-r-- 1 list list 8244 Sep 18 01:07 public/balug-test/2017/000004.html
> -rw-rw-r-- 1 list list 2553 Sep 18 01:07 public/balug-test/2017/000009.html
> -rw-rw-r-- 1 list list 7865 Sep 18 01:07 public/balug-test/2017/000006.html
> -rw-rw-r-- 1 list list 3443 Sep 18 01:07 public/balug-test/2017/subject.html
> -rw-rw-r-- 1 list list 2416 Sep 18 01:07 public/balug-test/2017/000011.html
> -rw-rw-r-- 1 list list 3390 Sep 18 01:07 public/balug-test/2017/000012.html
> -rw-rw-r-- 1 list list 3994 Sep 18 01:07 public/balug-test/2017/thread.html
> -rw-rw-r-- 1 list list 3075 Sep 18 01:07 public/balug-test/2017/000013.html
> -rw-rw-r-- 1 list list 3445 Sep 18 01:07 public/balug-test/2017/author.html
> -rw-rw---- 1 list list 1942 Sep 18 01:07  
> public/balug-test/database/2017-thread
> -rw-rw---- 1 list list 19497 Sep 18 01:07  
> public/balug-test/database/2017-article
> -rw-rw---- 1 list list 2455 Sep 18 01:07  
> public/balug-test/database/2017-subject
> -rw-rw---- 1 list list 1756 Sep 18 01:07 public/balug-test/database/2017-date
> -rw-rw---- 1 list list 2014 Sep 18 01:07  
> public/balug-test/database/2017-author
> -rw-rw---- 1 list list 847 Sep 18 01:07 public/balug-test/pipermail.pck
> $
> //That would, at least at quick glance, appear quite reasonable, and
> //seems to function okay, etc.
> //Is it consistent with the other archives (notably in
> //permissions/ownerships)?
> $ (cd /var/lib/mailman/archives/ && find */balug-announce -follow  
> -name database -type d -prune -o -type f -exec ls -ld \{\} \;)
> -rw-rw-r-- 1 list list 1204 Aug 14 01:10 private/balug-announce/index.html
> -rw-rw-r-- 1 list list 7800 Aug 14 01:10  
> private/balug-announce/2017-August.txt
> -rw-rw-r-- 1 list list 3552 Aug 14 03:27  
> private/balug-announce/2017-August.txt.gz
> -rw-rw---- 1 list list 856 Aug 14 01:10 private/balug-announce/pipermail.pck
> -rw-rw-r-- 1 list list 10651 Aug 14 01:10  
> private/balug-announce/2017-August/000000.html
> lrwxrwxrwx 1 list list 11 Aug 14 01:10  
> private/balug-announce/2017-August/index.html -> thread.html
> -rw-rw-r-- 1 list list 1751 Aug 14 01:10  
> private/balug-announce/2017-August/date.html
> -rw-rw-r-- 1 list list 1745 Aug 14 01:10  
> private/balug-announce/2017-August/subject.html
> -rw-rw-r-- 1 list list 1772 Aug 14 01:10  
> private/balug-announce/2017-August/thread.html
> -rw-rw-r-- 1 list list 1747 Aug 14 01:10  
> private/balug-announce/2017-August/author.html
> -rw-rw-r-- 1 list list 1204 Aug 14 01:10 public/balug-announce/index.html
> -rw-rw-r-- 1 list list 7800 Aug 14 01:10  
> public/balug-announce/2017-August.txt
> -rw-rw-r-- 1 list list 3552 Aug 14 03:27  
> public/balug-announce/2017-August.txt.gz
> -rw-rw---- 1 list list 856 Aug 14 01:10 public/balug-announce/pipermail.pck
> -rw-rw-r-- 1 list list 10651 Aug 14 01:10  
> public/balug-announce/2017-August/000000.html
> lrwxrwxrwx 1 list list 11 Aug 14 01:10  
> public/balug-announce/2017-August/index.html -> thread.html
> -rw-rw-r-- 1 list list 1751 Aug 14 01:10  
> public/balug-announce/2017-August/date.html
> -rw-rw-r-- 1 list list 1745 Aug 14 01:10  
> public/balug-announce/2017-August/subject.html
> -rw-rw-r-- 1 list list 1772 Aug 14 01:10  
> public/balug-announce/2017-August/thread.html
> -rw-rw-r-- 1 list list 1747 Aug 14 01:10  
> public/balug-announce/2017-August/author.html
> $
> //I'd say that all looks quite excellently consistent.  Hopefully our
> //umask value didn't even matter to the arch command, and it simply "did
> //the right thing" anyway ... but we set a prudent value - in case arch
> //didn't set it or explicity set permissions.  Maybe I'll (further) test
> //that at some point ... but that's lower on the priority queue.
> //And ... let's also test delete/add of a message to make sure that
> //works as we hope and expect ...
> //and that all tests good, no surprises there.
> //So ... onto merging into the archives the older raw mbox
> //bits obtained from DreamHost.com.