Hooray! I have switched most of my tools over, taking care to put
long-running queries (including PetScan queries) on the analytics server.
Initial tests see a ~5x improvement in query time!
On Tue, Sep 26, 2017 at 4:25 AM Dan Andreescu
wrote:
> In case you are tempted
Sorry to see you go, happy you'll return! :-)
On Mon, Mar 20, 2017 at 7:29 PM Pine W wrote:
> :( Sorry to read this. I look forward to "seeing" you when you get back
> from your break, and I hope that you have a nice vacation.
>
> Pine
>
>
> On Mon, Mar 20, 2017 at 12:23
Many thanks for finding and fixing it!
On Thu, Feb 9, 2017 at 4:50 PM Bryan Davis wrote:
> Magnus ran into a problem with one of his tools [0] that may affect
> others as well.
>
> The version of Node used for WMF's production services running on
> Debian Jessie has been
Hi,
I used to have this all the time, but it effectively stopped since the k8s
webservice.
At https://tools.wmflabs.org/giftbot/ I just see a directory listing, which
means the webservice is running, just not set up.
Is it the webservice that fails, or the bot part of giftbot?
On Thu, Jan 12,
Hi all,
is there an automatic backup of tools databases on Labs? I broke the
mix'n'match DB by mistake, would really like a DB restore or dump from
before noon 2016-12-26 (yesterday)!
___
Labs-l mailing list
Labs-l@lists.wikimedia.org
e nice to establish a norm now on naming :)
>
> Cheers,
>
> Chase
>
> On Wed, Nov 30, 2016 at 6:31 AM, Andre Klapper <aklap...@wikimedia.org>
> wrote:
>
> Hi Magnus,
>
> On Wed, 2016-11-30 at 11:51 +, Magnus Manske wrote:
> > So I now have a d
So I now have a diffusion repo for my quickstatements (V2) tool:
https://phabricator.wikimedia.org/diffusion/2010/
How can I attach a Phabricator task/issue to it? Neither "quickstatements"
nor "tool-quickstatements" show up in tags when I try to submit a task.
Likewise, how can I get a list for
28 odesílatel Alex Monk <am...@wikimedia.org> napsal:
>
> Try login.tools.wmflabs.org instead of tools-login.wmflabs.org ?
>
> On 9 November 2016 at 13:05, Magnus Manske <magnusman...@googlemail.com>
> wrote:
>
> When I try to ssh into tools-login.wmflabs.org or basti
When I try to ssh into tools-login.wmflabs.org or bastion, login just stops.
ssh -vvv stops at
debug2: fd 3 setting O_NONBLOCK
I didn't change anything on my side, so...
Help?
___
Labs-l mailing list
Labs-l@lists.wikimedia.org
I was thinking about trying out some React on Tools Labs, maybe write and
publish some components to be re-used in other tools or by third-party
developers.
Questions:
* Am I allowed to use React on Tools Labs? The license is not OSI-approved,
though AFAIK the Facebook addendum to the BSD license
t; > On Thu, Apr 28, 2016 at 7:11 PM, Magnus Manske
> > <magnusman...@googlemail.com> wrote:
> >> Maybe because portgrabber is gone?
> >>
> >> On Thu, Apr 28, 2016 at 2:39 PM Magnus Manske <
> magnusman...@googlemail.com>
> >> wrote:
> >
Maybe because portgrabber is gone?
On Thu, Apr 28, 2016 at 2:39 PM Magnus Manske <magnusman...@googlemail.com>
wrote:
> https://tools.wmflabs.org/autodesc/ is 502, even after webservice
> restart. Sometimes it's 404, though...
> node.js, maybe that's the issue.
>
> On Thu,
https://tools.wmflabs.org/autodesc/ is 502, even after webservice restart.
Sometimes it's 404, though...
node.js, maybe that's the issue.
On Thu, Apr 28, 2016 at 11:29 AM Yuvi Panda wrote:
> This has been successfully completed now - let me know if you run into
> any
Looks good now, thanks!
On Mon, Apr 25, 2016 at 5:47 PM Yuvi Panda wrote:
> Hey Magnus!
>
> Can you try again now? There was a small DNS outage around that time
> (https://phabricator.wikimedia.org/T124680) and could probably be
> related.
>
>
Can't restart the webservice for my "sighting" tool. I notice "wdq2sparql"
is also 502.
Error message:
"Timeout: could not stop job in 30s"
___
Labs-l mailing list
Labs-l@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/labs-l
Yay, and welcome to the Mountains of Madness :-)
On Fri, Apr 15, 2016 at 7:14 PM Mukunda Modell
wrote:
> On Fri, Apr 15, 2016 at 12:41 PM, Arthur Richards >
> wrote:
>
> > I for one am psyched to see this news :)
> >
> >
> Arthur: I am confident
Huh. WDQ seems to be the biggest problem, but I can't find the data - only
~20GB in total. Unless Yuvi has cleaned up in the meantime, or I don't have
permissions...
On Fri, Mar 4, 2016 at 7:45 PM Brion Vibber wrote:
> Thanks for the reminder ping! I still need
Ah, didn't see this, created a bug report:
https://phabricator.wikimedia.org/T119382
On Mon, Nov 23, 2015 at 12:30 PM Yetkin Sakal wrote:
> That is good news for us all, jcrespo.
>
> Cheers!
>
>
>
> On Monday, November 23, 2015 2:00 PM, Jaime Crespo
imedia.org/T119315> Expect some downtime in order
> to fix the issues.
>
> On Mon, Nov 23, 2015 at 2:01 PM, Magnus Manske <
> magnusman...@googlemail.com> wrote:
>
>> Ah, didn't see this, created a bug report:
>> https://phabricator.wikimedia.org/T119382
>>
"PHP_FCGI_MAX_REQUESTS" => "500"
> ),
> "bin-copy-environment" => (
> "PATH", "SHELL", "USER"
> ),
> "broken-scr
How can I change the upload_max_filesize and post_max_size values (usually
in php.ini) for a tool (using lighttpd)?
___
Labs-l mailing list
Labs-l@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/labs-l
I think Hay has had some issues with the auto-update. The crawler logs are
404, so we can't see when the last run happened.
On Fri, Aug 14, 2015 at 4:22 AM Magog The Ogre magog.the.o...@gmail.com
wrote:
Thanks all. But my tool information still isn't listed after I fixed my
toolinfo.json a few
Delete the tools-login line(s) from your .ssh/known_hosts and it should
work again
On Mon, Mar 23, 2015 at 12:47 PM Marc Miquel marcmiq...@gmail.com wrote:
Hello,
I've been trying to mount my tool labs space as an sftp disk as I do
normally when I want to code. This time I haven't been able
Can I setup a node.js webserver as a tool on Labs (shared, no dedicated
VM), and access it under the tool's URL? If so, any documentation as to how?
Cheers,
Magnus
___
Labs-l mailing list
Labs-l@lists.wikimedia.org
Thanks all. Looking forward to that webservice2 :-)
In the meantime, the doc says to use portgrabber, and another page points
to /usr/local/bin/portgrabber but that doesn't exist? Anyone know where it
went?
On Wed Feb 04 2015 at 6:57:08 PM Gabriel Wicke gwi...@wikimedia.org wrote:
On Wed, Feb
Yes, except the webservice I started there is unkillable (for me), so the
Wikidata game is now offline unless someone kills job 6027755 and restarts
a webservice that actually works.
On Thu, Nov 27, 2014 at 10:34 PM, Yuvi Panda yuvipa...@gmail.com wrote:
Trouble!
The trusty instance VM seems
suggests they are cold files in
the database, but I don’t understand how that should be happening with
indexes in that area.
Cyberpower678
English Wikipedia Account Creation Team
Mailing List Moderator
On Nov 26, 2014, at 09:54, Magnus Manske magnusman...@googlemail.com
wrote:
I already
Thank you all for the updates and the frantic work!
On Thu, Nov 6, 2014 at 6:55 PM, Andrew Bogott abog...@wikimedia.org wrote:
On 11/6/14 11:45 AM, Marc A. Pelletier wrote:
Hello all,
There is currently a hardware issue that took down labs. We're hard at
work on it and will give news the
: [Labs-l] Google bot
On 10/19/2014 03:50 PM, Magnus Manske wrote:
I vaguely remember that indexing bots (like the Google one) were
filtered out by Labs already?
They were, for some time, but then I got some fairly vehement
protestations that tools being unindexed by Google was a problem
Hi all,
one of my scripts needs to check on categories in many languages, for a
single script call. Right now, it's disabled because it drove runtime up
massively.
So, in essence , it goes through a list of languages, often 100. It opens
a DB connection on Labs to the respective Wikipedia, does
Hi,
I saw a high load (dozens of queries [1]) hitting one of my tools
(catscan2). The queries looked like they came from a template on French
Wikipedia (category name different, other parameters the same). Access log
shows (among other things) Google bot. When I added that to my bot
exclusion
https://tools.wmflabs.org/magnustools/toolstatus.html
On Fri, Oct 10, 2014 at 1:02 AM, Dmitry Trofimovich killwind...@yandex.ru
wrote:
Hello!
Catscan2 worked badly recent 3 weeks. It retrieves lists with a big delay
and if you set a deeper depth (4 and more) it can't handle request.
It
Going slightly on a tangent here:
For tools generally considered cool/useful, but where the maintainer does
not know about/want to bother with the tool directory, should we set up a
wiki page, add the JSON for these tools, and add the action=raw URL?
That would allow the community (as in, some of
,
description : An API for Wikidata items and properties.,
url : http://wdq.wmflabs.org/;,
keywords : wikidata, api, query,
author : Magnus Manske
}
If you have a web-hosted tool, simply stick it in the root of your
tools directory so that it's reachable
So wikidatawiki.labsdb has just gone away. Related?
On Tue, Aug 5, 2014 at 3:07 AM, Sean Pringle sprin...@wikimedia.org wrote:
2014-08-07 00:02:00 the third and final labsdb1003 will switch to MariaDB
10. There is a larger chunk of user data to migrate on this instance, so
the downtime will
Thanks!
On Tue, Aug 5, 2014 at 2:24 PM, Sean Pringle sprin...@wikimedia.org wrote:
On Tue, Aug 5, 2014 at 10:00 PM, Magnus Manske
magnusman...@googlemail.com wrote:
So wikidatawiki.labsdb has just gone away. Related?
No, https://bugzilla.wikimedia.org/show_bug.cgi?id=69144
I've been manually restarting about a dozen webservices for my tools in the
last 24h.
And before you say it, some of those were Hedonil's hand-rolled webservice.
Could we PLEASE either have a Labs-official, auto- and self-restarting
webservice, or something a little more stable than lighttpd (or
, 2014 at 5:34 PM, Tim Landscheidt
t...@tim-landscheidt.de wrote:
Magnus Manske magnusman...@googlemail.com wrote:
I've been manually restarting about a dozen webservices for my tools
in the
last 24h.
And before you say it, some of those were Hedonil's hand-rolled
webservice
As the maintainer of several dozen tools, this happens on a regular basis.
No automatic notification, nor automatic restart. Pitiful, really.
Hedonil has written a set of scripts to run the webservice in a more
reliable manner, and even has an auto-restarter, which I use for some of
the tools
/. This lists
Magnus Manske for wikidata-game, so now you'd need to know
that Magnus adds a link to his source repositories in prac-
tically every one of his tools, except this one :-). If you
look at those other tools, you get pointed to
https://bitbucket.org/magnusmanske where wikidata-game
Added, but it doesn't appear to be structured the same way as the other
Wikisource projects. Pretty but not-machine-translatable script doesn't
help with debugging.
On Wed, Apr 30, 2014 at 6:10 PM, Jayanta Nath jayanta...@gmail.com wrote:
Hi All,
Can it be possible to add Bengali Wikisource
+1M...
On Wed, Apr 23, 2014 at 9:09 AM, Hay (Husky) hus...@gmail.com wrote:
+1. It's really annoying to explain to end users that they should put
a trailing slash after a tool's URL. It doesn't make any sense. Please
guys, fix this.
-- Hay
On Wed, Apr 23, 2014 at 5:03 AM, Hedonil
If you are an admin for that machine, you can set the web address yourself
at [[Special:NovaProxy]] on wikitech.
Cheers,
Magnus
On Fri, Mar 28, 2014 at 8:21 AM, Thomas Gries m...@tgries.de wrote:
What are the *web addresses* to access the migrated instances for
example [1]?
Do I have to
On Mon, Mar 24, 2014 at 9:13 PM, Marc A. Pelletier m...@uberbox.org wrote:
It will, by offloading that work to tools-sumbit.
Funny - writing HTML forms, I too type sumbit *every. single. time.*
___
Labs-l mailing list
Labs-l@lists.wikimedia.org
The directory is renamed, not deleted.
Did you run finish-migration on eqiad?
On Wed, Mar 5, 2014 at 5:50 PM, Bryan White bgwh...@gmail.com wrote:
The public_html directory was deleted by the migrate-tool script
before the move. The directory is not present in equid after the move.
Is it
Calm down. I'm trying that migration command on 40-ish tools now; so far so
good, I'll let you know how it works out ;-)
On Tue, Mar 4, 2014 at 8:18 PM, Maarten Dammers maar...@mdammers.nl wrote:
Hi Marc,
Marc A. Pelletier schreef op 4-3-2014 19:40:
For roughly the following two weeks
Pro: Seems to attempt to copy the databases, if any, as well
Con: Seems to mysqldump them to STDOUT and then freeze up...
On Tue, Mar 4, 2014 at 8:25 PM, Marc-André Pelletier
mpellet...@wikimedia.org wrote:
On 03/04/2014 03:18 PM, Maarten Dammers wrote:
This is just ridiculous.
That will
, Marc A. Pelletier m...@uberbox.org wrote:
On 03/04/2014 03:27 PM, Magnus Manske wrote:
Pro: Seems to attempt to copy the databases, if any, as well
Con: Seems to mysqldump them to STDOUT and then freeze up...
D'oh! My fault, I broke a script.
-- Marc
Yup, NFS went bad again. Since the Foundation considers Labs low
priority, it'll probably stay down at least until Monday.
On Sun, Feb 16, 2014 at 12:35 PM, Russell Blau russb...@imapmail.orgwrote:
I can't ssh to either tools-login or tools-dev right now; I was able to
ssh onto
Labs is using lighttpd, so maybe this helps:
http://peyotest.blogspot.co.uk/2009/12/lighttpd-php-file-upload.html
https://rikkoss.wordpress.com/2013/07/05/change-max-upload-size-wordpress-lighttpd/
Cheers,
Magnus
On Thu, Feb 6, 2014 at 9:29 PM, Prolineserver prolineser...@gmail.comwrote:
Hi,
And - fixed (geohack).
On Fri, Jan 10, 2014 at 4:57 PM, Marc A. Pelletier m...@uberbox.org wrote:
On 01/10/2014 10:35 AM, Brad Jorsch (Anomie) wrote:
Is it allowed to hotlink to project sites, such as Commons or enwiki?
Yes. Those actually scale better than Labs ever could. :-)
--
It's a big ol' honking tool all right, which is why I made [1], which is
/much/ faster. However, some complex queries from catscan2, especially when
used in combination, are rather hard to do is a single query, and holding
everything in PHP memory is probably less efficient than temporary MySQL
I vaguely remember that there was a mirror of the pageview files on Tools
Labs, but I can't find the path. Is that still around, and where?
Note: It's /not/ this:
/shared/viewstats/
Thanks,
Magnus
___
Labs-l mailing list
Labs-l@lists.wikimedia.org
Would this suffice?
http://208.80.153.172/
On Mon, Oct 28, 2013 at 11:59 AM, Tanvi Gupta tanvi.gup...@gmail.comwrote:
Dear Admin,
I would like to create a new project in wikilabs with the name - Wikidata
Property-Item Identification with a motive to retrieve all the possible
items for
then that would be great support
to my project.
Thanks.
On Mon, Oct 28, 2013 at 7:19 PM, Magnus Manske
magnusman...@googlemail.com wrote:
Would this suffice?
http://208.80.153.172/
On Mon, Oct 28, 2013 at 11:59 AM, Tanvi Gupta tanvi.gup...@gmail.comwrote:
Dear Admin,
I would like to create
should be correct according to:
https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/Help#Naming_conventions
Cheers,
Magnus
On Thu, Oct 10, 2013 at 1:36 PM, Marc-André Pelletier
mpellet...@wikimedia.org wrote:
On 10/08/2013 06:42 PM, Magnus Manske wrote:
Is it possible that whatever caused
mpellet...@wikimedia.org wrote:
On 10/10/2013 09:03 AM, Magnus Manske wrote:
Interesting! I have added the file to one of my tools (book2scroll), and
it works, but how can I tell which server it uses?
You can see what webservers are started on the tool status page:
https://tools.wmflabs.org
:03 AM, Magnus Manske wrote:
Which should be correct according to:
https://wikitech.wikimedia.org/wiki/Nova_Resource:Tools/Help#Naming_conventions
It is, that was my fault -- that ugly hack with the names wasn't yet
propagated to webgrid-01. Should work now.
-- Marc
Yes! All up-and-running!
I shall now add .lighttpd.conf to - EVERYTHING! :-)
Thanks a lot,
Magnus
On Thu, Oct 10, 2013 at 4:00 PM, Marc-André Pelletier
mpellet...@wikimedia.org wrote:
On 10/10/2013 10:02 AM, Magnus Manske wrote:
Maybe the same issue I mentioned already, might have nothing
mpellet...@wikimedia.org wrote:
On 10/10/2013 11:04 AM, Magnus Manske wrote:
I shall now add .lighttpd.conf to - EVERYTHING! :-)
Oh, poop. I just noticed a bug. Please stand by. :-)
-- Marc
___
Labs-l mailing list
Labs-l@lists.wikimedia.org
Thanks, looks all good!
On Thu, Oct 10, 2013 at 5:47 PM, Marc A. Pelletier m...@uberbox.org wrote:
On 10/10/2013 11:20 AM, Magnus Manske wrote:
While you fix it, two things I noticed:
The problem was a nasty race condition with lockfiles caused by the
slightly-off NFS creation semantics
Odd, that was actually where I added a timeout:
$ctx = stream_context_create(array('http' = array('timeout' = 5)));
$file = file_get_contents ( $url , 0 , $ctx ) ;
That should limit the request to 5 sec max.
I have now deactivated the Template:Motd part completely. Can you chase the
dragon
And it's down again...
On Fri, Oct 4, 2013 at 3:16 PM, Mr. Maximilian Doerr
cybernet...@yahoo.comwrote:
It's working now. But it wasn't when you sent out that email.
On Oct 4, 2013, at 9:27 AM, Marc-André Pelletier mpellet...@wikimedia.org
wrote:
On 10/04/2013 09:15 AM, Mr. Maximilian
So again, for catscan2, which resources? Memory? DB connections? http
connections? I can see how catscan2 might be heavy, but catfood is rather
light-weight, and I really can't see how it could eat up http connections.
Is it possible that whatever caused PHP to melt down last time is really a
PHP
I have added some PHP incantations that /should/ be unnecessary but might
solve the connection problems.
Could you unblock catfood and see if it works better now?
On Tue, Oct 8, 2013 at 3:51 PM, Marc-André Pelletier
mpellet...@wikimedia.org wrote:
On 10/08/2013 10:27 AM, Magnus Manske wrote
Getting 500 error on all my tools...
___
Labs-l mailing list
Labs-l@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/labs-l
It has begun:
https://meta.wikimedia.org/wiki/User_talk:Erwin#Labs_port
Anyone else interested, please help me porting/maintaining; grunt work is
done ;-)
On Fri, Sep 27, 2013 at 11:22 AM, Silke Meyer silke.me...@wikimedia.dewrote:
Erwin85 is looking for people to maintain his very popular
There was a recent mail saying that Labs is not considered production
stability. Mainly a disagreement about how many 9s in the 99.9% that
represents.
On Wed, Sep 11, 2013 at 4:09 PM, Petr Bena benap...@gmail.com wrote:
Hi,
can you please explain to me from where is this information:
inode dates April 16, as does the broken cgi-bin,
which could indicate an earlier date of (non-visible?) corruption. The tool
was originally created March 20.
Cheers,
Magnus
On Wed, Aug 14, 2013 at 3:12 PM, Marc A. Pelletier m...@uberbox.org wrote:
On 08/12/2013 11:08 AM, Magnus Manske wrote
that it is in! (I think)
On 12 August 2013 17:08, Magnus Manske magnusman...@googlemail.comwrote:
Tried IRC, but no joy.
Labs user russblau has replaced the public_html directory of my tool
/data/project/geograph2commons
with ... some file that I (and the tool user) can't access. Random
Can someone notify russblau please?
On Mon, Aug 12, 2013 at 4:22 PM, Magnus Manske
magnusman...@googlemail.comwrote:
local-geograph2commons@tools-login:~$ pwd
/data/project/geograph2commons
local-geograph2commons@tools-login:~$ ls -l
total 268
-rw-r- 1 root local
2013 17:23, Magnus Manske magnusman...@googlemail.comwrote:
Can someone notify russblau please?
On Mon, Aug 12, 2013 at 4:22 PM, Magnus Manske
magnusman...@googlemail.com wrote:
local-geograph2commons@tools-login:~$ pwd
/data/project/geograph2commons
local-geograph2commons@tools-login:~$ ls
2013 17:08, Magnus Manske magnusman...@googlemail.comwrote:
Tried IRC, but no joy.
Labs user russblau has replaced the public_html directory of my tool
/data/project/geograph2commons
with ... some file that I (and the tool user) can't access. Random people
nuking my tools, WTF
Thanks Russ. Well, let's see if there are any roots on this list...
On Mon, Aug 12, 2013 at 4:58 PM, Russell Blau russb...@imapmail.org wrote:
Magnus Manske wrote:
Tried IRC, but no joy.
Labs user russblau has replaced the public_html directory of my tool
/data/project
Thanks Tim!
Now that I know that there's a good copy, I'm happy to delay restoration
a day or two if it helps to find the cause, and hopefully prevent it from
happening again.
Cheers,
Magnus
On Mon, Aug 12, 2013 at 7:42 PM, Tim Landscheidt t...@tim-landscheidt.dewrote:
Magnus Manske
Yay! Thanks!
On Fri, Jun 28, 2013 at 9:40 PM, Yuvi Panda yuvipa...@gmail.com wrote:
Redis Libraries for PHP(https://github.com/nicolasff/phpredis) and
Python(https://github.com/andymccurdy/redis-py) have now been
installed :)
--
Yuvi Panda T
http://yuvi.in/blog
to be finished, I
would happily use gerrit as well :P
On Mon, Jun 24, 2013 at 8:21 PM, Ryan Lane rlan...@gmail.com wrote:
On Mon, Jun 24, 2013 at 11:18 AM, Magnus Manske
magnusman...@googlemail.com wrote:
After a slight cp accident ;-) I was wondering if there is an easy way
to
set up
. That should also
prevent a recurrence in the interim before the the server restarts occur.
And at least the s5 replica caught up very quickly.
-Asher
On Thu, Jun 13, 2013 at 1:22 AM, Magnus Manske
magnusman...@googlemail.com wrote:
Update: According to select max(rev_timestamp) from
Update: According to select max(rev_timestamp) from revision, the replag
is ~15h. Any particular reason?
On Thu, Jun 13, 2013 at 8:53 AM, Magnus Manske
magnusman...@googlemail.comwrote:
At least wikidatawiki_p seems to have a replag in minutes or more. Is
there a way to see this, either a web
Thanks, it works!
Also, couldn't find bugtracker link for labs, so I'll post it here: Table
globalimagelinks is not in the commonswiki_p replica.
Thanks,
Magnus
On Mon, May 27, 2013 at 12:08 AM, Tim Landscheidt
t...@tim-landscheidt.dewrote:
Magnus Manske magnusman...@googlemail.com wrote
On Mon, Apr 22, 2013 at 6:58 PM, Ryan Lane rlan...@gmail.com wrote:
Btw, on your project request, is this something you believe will become
part of wikidata, or something that will live as a tool from now on? If
it's just going to be a tool, it should likely go into the tools project.
I
81 matches
Mail list logo