User talk:DamianZaremba/Archives/2014/April

WMF might've broken CBNG

According to a WMF staff member, the foundation reset session tokens for all users as a precaution against the Heartbleed bug (perm link). While I understand it was for user security, it also had the unintended side-effect of disabling a number of bots on Wikipedia, ClueBot NG being one of them. The bot is up, as the CBNG queue in STiki is able to catch recent edits, bot just can't revert what it sees.

What does baffle me though is how, while many bots were killed, such as HBC AIV helperbot5 and 7, were also killed, other bots, such as SineBot and AnomieBOT, resumed their normal duties as if nothing happened. --k6ka (talk | contribs) 02:07, 10 April 2014 (UTC)

I take that bit about the other bots being killed back. The AIV helperbots aren't editing because there isn't any work for them to do. --k6ka (talk | contribs) 02:09, 10 April 2014 (UTC)

Thanks for the poke - killing the session may have stopped the bot making reverts (re-login logic is a bit broken, since it just uses a cookie jar for everything after login on startup), I've now restarted the bot (sorry, was out last night) and it still seems to be having issues making reverts. Would resetting the sessions make registration time for users (non anon) disappear from the database? I'll try and sort the current issue out during the day, but it might have to wait until I get in from work.

Seeing things like this happening at the moment, but I think it's intermittent depending on which user is making the edits (so the bot might start reverting):

Processing: [[Manthana]]  http://en.wikipedia.org/w/index.php?diff=603568633&oldid=603568610 * Paradoxs * (+4) /* About the series */
Forked - 11459

10/04/2014 08:24:23
array(13) {
  ["namespace"]=>
  string(5) "Main:"
  ["namespaceid"]=>
  int(0)
  ["title"]=>
  string(8) "Manthana"
  ["flags"]=>
  string(0) ""
  ["url"]=>
  string(66) "http://en.wikipedia.org/w/index.php?diff=603568633&oldid=603568610"
  ["revid"]=>
  string(9) "603568633"
  ["old_revid"]=>
  string(9) "603568610"
  ["user"]=>
  string(8) "Paradoxs"
  ["length"]=>
  string(2) "+4"
  ["comment"]=>
  string(22) "/* About the series */"
  ["timestamp"]=>
  int(1397118263)
  ["line"]=>
  string(121) "[[Manthana]]  http://en.wikipedia.org/w/index.php?diff=603568633&oldid=603568610 * Paradoxs * (+4) /* About the series */"
  ["rawline"]=>
  string(153) "^C14[[^C07Manthana^C14]]^C4 ^C10 ^C02http://en.wikipedia.org/w/index.php?diff=603568633&oldid=603568610^C ^C5*^C ^C03Paradoxs^C ^C5*^C (+4) ^C10/* About the series */^C"
}
array(5) {
  ["common"]=>
  array(4) {
    ["creator"]=>
    string(6) "Kali-K"
    ["page_made_time"]=>
    int(1192871449)
    ["num_recent_edits"]=>
    string(2) "48"
    ["num_recent_reversions"]=>
    string(1) "0"
  }
  ["user_reg_time"]=>
  NULL
  ["user_warns"]=>
  string(1) "0"
  ["user_edit_count"]=>
  string(2) "49"
  ["user_distinct_pages"]=>
  string(2) "28"
}
API error.

Notice there is no registration time... and the value in the database is broken (that array is built from a few database queries):

tools.cluebot@tools-login:~$ mysql --defaults-file=~/replica.my.cnf -h enwiki.labsdb
Welcome to the MariaDB monitor.  Commands end with ; or \g.
Your MariaDB connection id is 30893381
Server version: 5.5.34-MariaDB-1~precise-log mariadb.org binary distribution

Copyright (c) 2000, 2014, Oracle, Monty Program Ab and others.

Type 'help;' or '\h' for help. Type '\c' to clear the current input statement.

MariaDB [(none)]> use enwiki_p
Reading table information for completion of table and column names
You can turn off this feature to get a quicker startup with -A

Database changed
MariaDB [enwiki_p]> SELECT `user_registration` FROM `user` WHERE `user_name` = "Paradoxs";
+-------------------+
| user_registration |
+-------------------+
| NULL              |
+-------------------+
1 row in set (0.00 sec)

Actually most of the data seems broken

MariaDB [enwiki_p]> SELECT * FROM `user` WHERE `user_name` = "Paradoxs";
+---------+-----------+----------------+---------------+------------------+------------+--------------+--------------+------------+--------------------------+------------------+--------------------------+-------------------+-------------------+----------------+
| user_id | user_name | user_real_name | user_password | user_newpassword | user_email | user_options | user_touched | user_token | user_email_authenticated | user_email_token | user_email_token_expires | user_registration | user_newpass_time | user_editcount |
+---------+-----------+----------------+---------------+------------------+------------+--------------+--------------+------------+--------------------------+------------------+--------------------------+-------------------+-------------------+----------------+
|  249589 | Paradoxs  |                |          NULL |             NULL |       NULL |         NULL |         NULL |       NULL |                     NULL |             NULL |                     NULL | NULL              |              NULL |             49 |
+---------+-----------+----------------+---------------+------------------+------------+--------------+--------------+------------+--------------------------+------------------+--------------------------+-------------------+-------------------+----------------+
1 row in set (0.01 sec)

If you have time to ask around/know more than I do (I really should spend more time doing wikipedia/media/labs stuff heh) feel free to call me an idiot :)

Edit: Maybe this is a thing

MariaDB [enwiki_p]> SELECT count(*) FROM `user` WHERE `user_registration` is NULL;
+----------+
| count(*) |
+----------+
|   490766 |
+----------+
1 row in set (7.47 sec)

MariaDB [enwiki_p]> SELECT count(*) FROM `user` WHERE `user_registration` is not NULL;
+----------+
| count(*) |
+----------+
| 20646374 |
+----------+
1 row in set (7.50 sec)

I could fall-back to the anon method (basically using the current time) if this is NULL I suppose, but I thought this field should always be set hmm.

- Damian Zaremba ^{(talk • contribs)} 08:33, 10 April 2014 (UTC)