IRC logs for #aegir, 2014-06-08 (GMT)

2014-06-07
2014-06-09
TimeNickMessage
[10:09:57]* jerryitt has joined #aegir
[11:16:41]* johnstorey has quit (Quit: Nodding off now.)
[11:43:45]* realityloop is now known as realityloopAFK
[11:44:22]* realityloopAFK is now known as realityloop
[12:30:55]* jerryitt has quit (Quit: Connection closed for inactivity)
[13:00:29]* Egyptian[Home]1 has quit (Ping timeout: 252 seconds)
[13:05:16]* ckosloff has quit (Quit: Konversation terminated!)
[13:28:13]* btopro has joined #aegir
[13:29:36]* btopro has left #aegir ()
[14:40:24]* realityloop is now known as realityloopAFK
[15:07:56]* johnstorey has joined #aegir
[15:15:28]* johnstorey has quit (Quit: joining all red shirts in their final fate)
[15:41:07]* Egyptian[Home] has joined #aegir
[16:05:09]* abyshakes has joined #aegir
[16:13:48]* christefano has joined #aegir
[17:34:34]* christefano has quit (Quit: christefano)
[17:44:18]* Egyptian[Home] has quit (Ping timeout: 240 seconds)
[18:20:09]* realityloopAFK is now known as realityloop
[18:21:19]* realityloop is now known as realityloopAFK
[18:30:29]* e-anima has joined #aegir
[19:03:32]* lolmaus has joined #aegir
[20:06:32]* darthsteven has joined #aegir
[20:07:52]* lolmaus has quit (Quit: No Ping reply in 180 seconds.)
[20:14:02]* lolmaus has joined #aegir
[20:14:23]* lolmaus has quit (Remote host closed the connection)
[20:17:27]* darthsteven has quit (Quit: darthsteven)
[21:10:46]* arosboro has joined #aegir
[21:58:10]* Egyptian[Home] has joined #aegir
[22:03:35]<arosboro>does the aegir team still do weekly scrum?
[22:06:34]* Egyptian[Home] has quit (Ping timeout: 265 seconds)
[22:06:40]* DecipheredAFK has quit (*.net *.split)
[22:14:26]* DecipheredAFK has joined #aegir
[22:15:23]* ivanjaros has joined #aegir
[22:35:46]* Egyptian[Home] has joined #aegir
[23:23:20]* ckosloff has joined #aegir
[23:23:22]* jerryitt has joined #aegir
[01:16:17]* joestewart has quit (*.net *.split)
[01:17:24]* joestewart has joined #aegir
[01:19:46]* johnstorey has joined #aegir
[01:30:55]* jerryitt has quit (Quit: Connection closed for inactivity)
[01:42:24]* johnstorey has quit (Quit: Nodding off now.)
[02:02:48]* darthsteven has joined #aegir
[02:05:09]* gandhiano has joined #aegir
[02:10:17]* gandhiano has quit (Ping timeout: 252 seconds)
[02:25:23]* chertzog has joined #aegir
[02:25:33]<chertzog>anyone around?
[02:26:43]* jerryitt has joined #aegir
[02:35:07]* darthsteven has quit (Quit: darthsteven)
[02:55:22]* christefano has joined #aegir
[02:56:08]<arosboro>hey chris
[02:56:11]<arosboro>my phone's dead
[02:56:34]<arosboro>@chertzog hi
[02:57:05]<chertzog>hey there arosboro
[02:57:24]<arosboro>not familiar with my mac irc client
[02:57:41]<chertzog>what are you using?
[02:57:45]<arosboro>xchat I think
[02:57:51]<chertzog>use limechat
[02:57:54]<arosboro>nice
[02:58:00]<arosboro>so I was looking at cloud files
[02:59:04]<arosboro>comes out to like $378/mo for 4TB of storage not including IO
[03:00:04]<arosboro>makes sense short term though
[03:00:06]<chertzog>sure. but how much space would we actually need
[03:00:28]<arosboro>roi will be quick when we need >1TB
[03:01:49]<arosboro>glusterfs scales linerarly if you do it right, last time I played with it, I wasn't imlpementing correctly
[03:02:05]<chertzog>ok
[03:02:54]<arosboro>but it works as an nfs/cifs client so it might be able to fit in with cluster
[03:03:57]<arosboro>nfs is recommended for small files and read performance
[03:04:38]<chertzog>ok
[03:04:50]<chertzog>im getting my own cluster set up now on my local box
[03:04:56]<arosboro>nice
[03:05:03]<arosboro>will be good for dev work
[03:05:12]<chertzog>got 5 vms set up. starting the galera stuff now
[03:05:32]<arosboro>did you find a good guide?
[03:05:40]<chertzog>i figured 2 webs and 2 dbs would be fine for dev
[03:05:49]<chertzog>right now im following https://blog.mariadb.org/installing-mariadb-galera-cluster-on-debian-ubu...
[03:06:26]<arosboro>I think it might make sense to test the cluster with one role and then add vms, so I wouldn't go crazy with the idea of scaling it up since in practice we know that works
[03:07:34]<arosboro>or I mean each service should work, and be trivial to configure without spending time on optimizing, but lets focus on the platform itself and contributing to 7.x-3.x-dev where we can at the moment
[03:08:07]<chertzog>yeah, that was my plan
[03:08:17]<chertzog>work out any bugs with the pack/cluster modules
[03:08:25]<arosboro>cool, still haven't figured out why I don't have a site alias
[03:08:35]<arosboro>for the sql server
[03:08:41]<arosboro>it's not in .drush
[03:08:49]<arosboro>or /var/aegir/config
[03:09:07]<arosboro>so I'm thinking something went wrong during provision of that server node
[03:09:19]<chertzog>let me get up to that point, and see what happens on my end
[03:09:35]<arosboro>nice good to have to environments to try this
[03:11:12]<gboudrias>helmo: ping
[03:12:02]<gboudrias>jonpugh: Hey
[03:22:27]<chertzog>arosboro: did you have any issue starting mysql on other nodes in the galera cluster
[03:22:53]<arosboro>not that I'm aware of
[03:24:02]<arosboro>did you stop them all then add --wresp-new-cluster for the first node?
[03:24:16]<arosboro>they were started by apt-get for me
[03:24:21]<chertzog>yeah, the first one started just fine
[03:24:33]<chertzog>trying to start the second one dies
[03:25:46]<arosboro>check that the configs are all the same except the line for the wresp address and the node's ip
[03:26:16]<arosboro>are you using vhost private networks or nat?
[03:26:32]<arosboro>er virtualbox private networks
[03:26:40]<chertzog>yep
[03:27:05]<chertzog>do you use the same wsrep_cluster_address list for all nodes?
[03:27:20]<arosboro>yes it should contain all nodes ips
[03:27:35]* realityloopAFK is now known as realityloop
[03:27:39]<arosboro>gcomm://first_ip,second_ip,third_ip
[03:27:48]<chertzog>the galera.cnf file is exactly the same for all ndoes correct
[03:28:53]<arosboro>except for wsrep_node_address and wsrep_node_name
[03:31:33]<arosboro>/etc/mysql/debian.cnf should contain maintenance conf that is the same on all servers as well
[03:33:01]<chertzog>did you have to edit debian.cnf at all
[03:34:14]<chertzog>HA got it
[03:34:37]<chertzog>the stupid tutorial didnt have the _node_address or node_name in the example
[03:41:24]<arosboro>chertzog, nice time to see if aegir lets you add a server
[03:43:18]<arosboro>nice macbook vpn lets me use the internet while i'm connected
[03:43:56]<arosboro>I'm doing my dev on the patio today, great day for it
[04:01:24]<chertzog>same error. no alias
[04:03:04]<arosboro>time to make an issue on d.o?
[04:03:27]<chertzog>give me a couple. i think i see the issue
[04:03:30]<ergonlogic>arosboro: hi, I got your error log via email. But you can always post it to a pastebin, and just link it in the issue
[04:03:46]<ergonlogic>chertzog: how've you been?
[04:03:50]<arosboro>ergonlogic: thanks
[04:04:07]<chertzog>not too bad. busy with client work. glad to finally have some time for fun again
[04:04:39]<chertzog>arosboro and I are working on some aegir d7 stuff together
[04:04:57]<arosboro>multiserver goodness
[04:05:58]<ergonlogic>nice
[04:06:54]<ergonlogic>arosboro: so, installing Aegir3 will work on Drush 7, at commit 8609fc29734ea
[04:06:59]<ergonlogic>but no later, at this point
[04:07:30]<ergonlogic>that's a commit on drush, btw
[04:09:08]<arosboro>ok, I noticed that in the queue. Would you prefer development and testing be done with drush 7 or drush6 at the moment?
[04:11:03]<arosboro>chertzog and I currently have two environments running with drush 6 aegir 3-dev
[04:13:19]<chertzog>arosboro: it looks like the hosting_name isnt being set on new server inserts in hosting_server_insert
[04:19:47]<ergonlogic>arosboro: feel free to use Drush 6
[04:20:19]<ergonlogic>I'll spend some time figuring out why that change to broke our server aliases
[04:20:52]<ergonlogic>we'll need aegir3 operating on drush7 before we release, for d8 support, but that's a little ways off yet
[04:21:07]<arosboro>ok
[04:21:47]<chertzog>yeah, doing a node_load on the server then a hosting_context_regsiter followed by a verify task creates the alias
[04:22:11]<chertzog>its something with the revision checking in hosting_server_insert
[04:22:18]<arosboro>ergonlogic: so one of the commits on Apr 08, by mosh broke the install scripts?
[04:22:30]* ivanjaros has quit (Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org)
[04:22:50]<arosboro>moshe*
[04:22:52]<ergonlogic>yeah, he moved where the --backend option gets set
[04:23:01]<ergonlogic>which seems pretty innocuous
[04:25:31]<arosboro>I'm not very familiar with the inner workings of drush yet, but is context basically a variable set in the environment during script execution?
[04:40:12]* mikebarkas has joined #aegir
[04:46:01]* abyshakes has quit (Remote host closed the connection)
[04:52:27]<arosboro>chertzog: you may want to try using aegir-up need a quick throw-away environment
[04:52:50]<chertzog>ive used that a few times before. its nice
[04:53:42]<arosboro>looks like dropbox hosts the wheezy image throtleing at 100 kbits :(
[04:54:42]<chertzog>im having another galera issue. the cluster is not recognizing the second db
[04:55:23]<arosboro>you mean from the site created by aeigr?
[04:55:36]<arosboro>or one node is down
[04:56:40]<chertzog>node 2 is running mysql but its not replicating
[04:57:16]<chertzog>node 2 is up and able to be queried etc, but its not being recognized
[04:57:20]<chertzog>by node 1
[05:01:58]<chertzog>got it. there were some rogue mysql processes, and it wouldnt propery restart
[05:03:32]<chertzog>@arosboro i was able to get a site installed via aegir on a 2 node galera cluster
[05:03:43]<chertzog>arosboro: ^
[05:03:51]<arosboro>cool
[05:04:29]<chertzog>now on to the web heads
[05:10:00]<arosboro>chertzog: brb gotta switch computers
[05:12:47]* arosboro has quit (Quit: This computer has gone to sleep)
[05:27:06]<chertzog>ergonlogic: can you explain the difference between the pack and cluster modules?
[05:27:18]<ergonlogic>I can try...
[05:27:59]<ergonlogic>pack will mount a common platform via nfs on webnodes
[05:29:15]<ergonlogic>whereas cluster, iirc, will rsync the platform out to the webnodes
[05:30:13]<chertzog>ok i think i get it
[05:31:24]* arosboro has joined #aegir
[05:43:27]* josh_k has joined #aegir
[05:49:39]* arosboro1 has joined #aegir
[05:49:57]* arosboro has quit (Quit: This computer has gone to sleep)
[05:50:00]<arosboro1>chertzog: yt?
[05:50:05]* arosboro1 is now known as arosboro
[05:50:11]<chertzog>whats up
[05:50:27]<arosboro>just configured my routes correctly can use internet and test the cluster
[05:50:35]<chertzog>nice.
[05:50:56]<arosboro>using pidgin much better than xchat aqua
[05:51:14]<chertzog>check out limewire
[05:51:17]<chertzog>limechat
[05:51:23]<arosboro>lol pirate!
[05:52:15]<arosboro>that's a mac thing xchat is supposed to be the best on ubuntu I think, that or b*tchx if you wanna go terminal :)
[05:52:59]<arosboro>chertzog: I just need tab support for user mentions
[05:56:16]<arosboro>ergonlogic: is there a mirror for the aegir up base image?
[06:02:01]* arosboro_ has joined #aegir
[06:02:22]<chertzog>arosboro: i just installed a site on a 2 node web cluster with a galera db cluster
[06:02:27]<chertzog>awesome!!!
[06:06:42]* josh_k has quit ()
[06:07:16]<arosboro>man the other computer is still on irc, and I get kicked out of my nic when the client comes back
[06:07:22]<arosboro>oh nvm
[06:07:31]<arosboro>chertzog: is there a patch so I can try :)
[06:07:41]<chertzog>for what?
[06:08:12]<arosboro>for the db issue
[06:08:33]<arosboro>chertzog, wait did you patch the issues with cluster suite as well?
[06:08:51]<chertzog>not yet. those didnt matter yet.
[06:09:18]<arosboro>so you have 6 servers in your cluster?
[06:09:33]<arosboro>or is it 7?
[06:09:37]<chertzog>5
[06:09:38]<hefring>http://drupal.org/node/5 => Support => 0 comments, 4 IRC mentions
[06:10:04]<chertzog>arosboro: just to to devel/php and run:
[06:10:24]<chertzog>$node = node_load(XXX);
[06:10:25]<chertzog> $hosting_name = isset($node->hosting_name) ? $node->hosting_name : 'server_' . preg_replace("/[!\W\.\-]/", "", $node->title);
[06:10:25]<chertzog> hosting_context_register($node->nid, $hosting_name );
[06:10:45]<chertzog>just load the server node you need
[06:10:51]<chertzog>then run a verify task on the server
[06:11:16]<chertzog>those are lines 433-434 of hosting_server.module
[06:11:53]<arosboro>did you create a pack?
[06:12:01]<chertzog>cluster not a pack
[06:16:18]* arosboro_ has quit (Quit: This computer has gone to sleep)
[06:19:39]<arosboro>nice got a warning Calling hook drush_site_data_post_provision_install[SITE_DATA] d()->site_data:Returned from hook drush_site_data_post_provision_install
[06:19:45]<arosboro>but overall looks good
[06:20:00]<chertzog>so you got a site installed?
[06:22:15]<arosboro>site installed locally, time to provision some slave apache servers
[06:27:24]* arosboro_ has joined #aegir
[06:30:55]* jerryitt has quit (Quit: Connection closed for inactivity)
[06:38:55]<arosboro>chertzog: how did you share platforms with the web servers? or is not load balanced?
[06:40:06]<chertzog>ok, so what you have to do is add each web server as a server node with apache. then add another server node that acts as just a cluster, and you pick which web servers you want to include
[06:40:31]<chertzog>then you need to edit the platform node and change the server to your newly created cluster
[06:40:41]<chertzog>and run verify tasks after everything
[06:44:46]<ergonlogic>arosboro: sorry, was on the road there...
[06:44:56]<ergonlogic>what do you need?
[06:47:40]<arosboro>ergonlogic: it's ok now, I was just waiting 30min to download the base image for aegir-up
[06:47:53]<ergonlogic>ah, ok
[06:48:05]<ergonlogic>I haven't touched it in a long time now
[06:48:19]<ergonlogic>I have some packer templates for aegir3 now
[06:48:36]<ergonlogic>so I'm working on making a regular build of that for testing/dev
[06:50:22]<arosboro>cool
[06:51:19]* arosboro_ has quit (Quit: This computer has gone to sleep)
[07:00:12]<arosboro>chertzog: do you select a the hostmaster from servers when you create a cluster?
[07:00:48]<arosboro>chertzog: nvm you explained it above
[07:04:06]<arosboro>is the cluster node tied to a specific host in addition to managing a list of servers in the cluster?
[07:04:38]<chertzog>now its time to add a varnish node in front of the cluster
[07:14:07]<arosboro>chertzog: looking at http://community.aegirproject.org/node/388 ?
[07:14:48]<chertzog>and http://www.lullabot.com/blog/article/configuring-varnish-high-availabili...
[07:48:14]<chertzog>and it works
[07:49:11]<arosboro>awesome
[07:49:56]<arosboro>I just made a cluster, resolved to an ip of some random server on the network, does that seem to matter, or is it actually supposed to exists?
[07:50:21]<arosboro>chertzog: ^
[07:50:40]<chertzog>it doesnt matter
[07:51:27]<chertzog>so i put a varnish server in front of the cluster that points at the 2 web heads of the cluster
[07:53:21]<arosboro>nice I have aegir-master and aegir-web-01 clustered together with 3 sql servers, not enough memory for varnish
[07:54:08]<chertzog>yeah, im maxing out my local machine with 6 vms running
[07:54:54]<chertzog>so it looks like clustering works just fine
[07:55:56]<chertzog>i think the nginx stuff needs some work though
[07:57:30]<arosboro>I get a error when I try to migrate
[07:57:41]<chertzog>what are you migrating
[07:58:39]* realityloop is now known as realityloopAFK
[07:59:44]<arosboro>the drupal7 site from this morning
[07:59:59]<chertzog>where are you trying to move it
[08:00:17]<chertzog>and what error are you getting
[08:00:19]<arosboro>intending to go to the cluster but it errors out as soon as I click migrat
[08:01:19]<arosboro>/var/aegir/hostmaster-7.x-3.x/profiles/hostmaster/modules/hosting/package/hosting_package.instance.inc:414 arg 2 passed to db_query must be array string given
[08:02:07]* e-anima has quit (Quit: reallife not found)
[08:07:47]<chertzog>looks like there is an error in this syntax
[08:07:51]<chertzog>CREATE TEMPORARY TABLE :name
[08:07:51]<chertzog> AS SELECT p.nid, p.short_name, p.old_short_name, i.version_code, i.schema_version, i.version, i.status
[08:07:51]<chertzog> FROM {hosting_package} p
[08:07:53]<chertzog> LEFT JOIN {hosting_package_instance} i
[08:07:55]<chertzog> ON p.nid = i.package_id
[08:07:57]<chertzog> WHERE i.rid = :ref and i.platform != :platform
[08:11:29]<ergonlogic>chertzog arosboro: I take you'll file bug reports whenever you come across issue on Aegir 3?
[08:11:39]<chertzog>yeah.
[08:12:00]<ergonlogic>cool, thanks
[08:12:04]<chertzog>though this one seems to be eluding me
[08:12:12]<chertzog>any ideas on that query ^^
[08:12:23]<ergonlogic>it looks like most of what you're doind is working though. That's quite encouraging
[08:12:33]<arosboro>yup
[08:12:37]<ergonlogic>what's that from?
[08:12:50]<chertzog>hosting_package.instance.inc
[08:12:53]<ergonlogic>oh hosting_package.instance.inc?
[08:12:53]<arosboro>chertzog:
[08:12:56]<ergonlogic>right
[08:12:58]<arosboro>it's argument 2 not one
[08:13:06]<arosboro>it needs to be an options array of replacement patterns
[08:13:12]<chertzog>its being called when clicking on migrate
[08:13:21]<arosboro>it's set to the temporary table name
[08:13:36]<chertzog>arosboro: yeah, i fixed that already. there is still something wrong with the query
[08:13:44]<chertzog> db_query("CREATE TEMPORARY TABLE :name
[08:13:44]<chertzog> AS SELECT p.nid, p.short_name, p.old_short_name, i.version_code, i.schema_version, i.version, i.status
[08:13:44]<chertzog> FROM hosting_package p
[08:13:45]<chertzog> LEFT JOIN hosting_package_instance i
[08:13:47]<chertzog> ON p.nid = i.package_id
[08:13:49]<chertzog> WHERE i.rid = :ref and i.platform != -1", // @todo remove magic number
[08:13:51]<chertzog> array(':name' => $names[$ref], ':ref' =>$ref));
[08:14:07]* johnstorey has joined #aegir
[08:15:11]* mikebarkas has quit ()
[08:16:33]<chertzog>we should probably try using https://api.drupal.org/api/drupal/includes!database!database.inc/functio...
[08:17:58]<ergonlogic>chertzog: what error are oyu seeing now?
[08:18:45]<ergonlogic>lso, db_query returns a db_query object, no?
[08:18:50]<ergonlogic>also*
[08:19:18]<ergonlogic>the result appears to be used in hosting_package_comparison()
[08:19:25]<arosboro>The name of the temporary table.
[08:20:04]<arosboro>oh you're right
[08:20:22]* realityloopAFK is now known as realityloop
[08:20:40]<ergonlogic>so, it is passing the $names[$ref]...
[08:22:00]<ergonlogic>try removing that magic number
[08:22:27]<ergonlogic>I think some of the re-factoring of the packages system might have changes that value
[08:22:43]<ergonlogic>soince we now track site-specific packages
[08:22:47]<ergonlogic>maybe...
[08:26:33]* johnstorey has quit (Quit: Nodding off now.)
[08:26:46]<chertzog>this is still erroring out:
[08:26:49]<chertzog> // Create the temporary table.
[08:26:49]<chertzog> db_query("CREATE TEMPORARY TABLE :name
[08:26:49]<chertzog> AS SELECT p.nid, p.short_name, p.old_short_name, i.version_code, i.schema_version, i.version, i.status
[08:26:50]<chertzog> FROM {hosting_package} p
[08:26:52]<chertzog> LEFT JOIN {hosting_package_instance} i
[08:26:54]<chertzog> ON p.nid = i.package_id
[08:26:56]<chertzog> WHERE i.rid = :rid and i.platform != :platform", // @todo remove magic number
[08:26:58]<chertzog> array(':name' => $names[$ref], ':rid' =>$ref, ':platform' => -1));
[08:27:23]<chertzog>PDOException: SQLSTATE[42000]: Syntax error or access violation: 1064 You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near ''_hosting_comparision_119' AS SELECT p.nid, p.short_name, p.old_sh' at line 1: CREATE TEMPORARY TABLE :name AS SELECT p.nid, p.short_name, p.old_short_name, i.version_code, i.schema_version, i.version, i.status FROM {hosti
[08:27:23]<chertzog>ng_package} p LEFT JOIN {hosting_package_instance} i ON p.nid = i.package_id WHERE i.rid = :rid and i.platform != :platform; Array ( [:name] => _hosting_comparision_119 [:rid] => 119 [:platform] => -1 ) in _hosting_package_temporary_table() (line 414 of
[08:27:46]<chertzog>PDOException: SQLSTATE[42000]: Syntax error or access violation: 1064 You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near ''_hosting_comparision_119' AS SELECT p.nid, p.short_name, p.old_sh' at line 1: CREATE TEMPORARY TABLE :name AS SELECT p.nid, p.short_name, p.old_short_name, i.version_code, i.schema_version, i.version, i.status FROM {hosti
[08:27:46]<chertzog>ng_package} p LEFT JOIN {hosting_package_instance} i ON p.nid = i.package_id WHERE i.rid = :rid and i.platform != :platform; Array ( [:name] => _hosting_comparision_119 [:rid] => 119 [:platform] => -1 ) in _hosting_package_temporary_table() (line 414 of
[08:29:14]* johnstorey has joined #aegir
[08:29:21]* gandhiano has joined #aegir
[08:29:27]* johnstorey has quit (Client Quit)
[08:30:20]<ergonlogic>I think arosboro's suggestion is sound: try db_query_temporary()
[08:30:53]<arosboro>actually chertzog found that
[08:31:24]<chertzog>yeah, i just replaced it with that, and its getting there. there is another bug though
[08:33:57]<ergonlogic>also, for the magic number, iirc, -1 is the platform id when the package is installed in the site
[08:34:16]<ergonlogic>so it's be better to use a constant
[08:34:30]<ergonlogic>since '-1' for a pid is kinda cryptic
[08:34:51]<ergonlogic>though that probably isn't the issue
[08:40:46]<chertzog>this is the next issue:
[08:40:49]<chertzog> $result = db_query("SELECT count(c.nid) AS error
[08:40:49]<chertzog> FROM :currenttable c
[08:40:49]<chertzog> LEFT JOIN :targettable t
[08:40:50]<chertzog> ON c.nid=t.nid
[08:40:52]<chertzog> WHERE (t.schema_version > 0) && (c.schema_version > t.schema_version)
[08:40:54]<chertzog> AND c.status = :cstatus",
[08:40:56]<chertzog> array(':currenttable' => $current_table,
[08:40:58]<chertzog> ':targettable' => $target_table,
[08:41:00]<chertzog> ':cstatus' => 1,
[08:41:02]<chertzog> )
[08:41:04]<chertzog> );
[08:41:25]<arosboro>is that hosting_package_comparison() (line 348
[08:41:35]<chertzog>yes
[08:41:44]<chertzog>the db_query_temporary works
[08:41:54]<arosboro>yeah I just re-factored as well
[08:42:01]<chertzog>but the comparison part is the next thing that is broken
[08:42:10]<chertzog>did you get it working?
[08:42:20]<arosboro>just got the same error you're on
[08:44:01]* gandhiano has quit (Ping timeout: 240 seconds)
[08:47:20]<chertzog>hmm. inserting the table name variables directly into the query and then running it comes up with :
[08:47:21]<chertzog>PDOException: SQLSTATE[42S02]: Base table or view not found: 1146 Table 'master.hosting_comparision_119' doesn't exist:
[08:47:32]<chertzog>so it looks like the temp table is staying around
[08:48:32]<arosboro>no that's not the temp table name
[08:48:50]<arosboro>it's looking up the wrong temp table using _hosting_comparision_' . trim($ref)
[08:50:22]<arosboro>should be something like db_temporary_1
[08:54:10]<arosboro>try changing static $names to:
[08:54:10]<arosboro>$names = &drupal_static(__FUNCTION__);
[08:54:11]<arosboro> if (!isset($names, $names[$ref])) {
[08:56:52]<arosboro>and then accessing the static with $names = &drupal_static('_hosting_package_temporary_table'); $names[$ref] == 'temporary_table_1'
[09:06:42]<arosboro>chertzog: nevermind, I didn't look at the query first I'm a little tired/distracted atm
[09:08:54]<arosboro>chertzog: I think I'm gonna call it a night...
[09:16:50]<chertzog>i figured it out: https://drupal.org/node/743090
[09:16:51]<hefring>https://drupal.org/node/743090 => Using Placeholders for database table names in db_* functions [#743090] => 0 comments, 1 IRC mention
[09:17:02]<chertzog>will submit patch later
[09:21:42]* johnstorey has joined #aegir
[09:33:54]* mozodev has joined #aegir
[09:45:04]* burgess has joined #aegir
[09:55:01]* johnstorey has quit (Quit: Nodding off now.)
[09:55:57]* johnstorey has joined #aegir
[09:56:19]<ergonlogic>chertzog: cool thanks, I'll take a look in a bit