| 00:21.47 | *** join/#uphpu dharmatech (~hayduke@166-70-57-177.ip.xmission.com) |
| 00:22.22 | *** part/#uphpu mgeary (~mgeary@166.70.44.49) |
| 00:57.03 | *** join/#uphpu jtsnow (~jtsnow@VDSL-151-118-130-166.DNVR.QWEST.NET) |
| 01:28.01 | *** join/#uphpu jtsnow (~jtsnow@VDSL-151-118-130-166.DNVR.QWEST.NET) |
| 03:18.21 | *** join/#uphpu jsmith (~jsmith@fedora/jsmith) |
| 03:18.21 | *** join/#uphpu pelaofeliz (~pelaofeli@67.137.236.66) |
| 03:18.21 | *** mode/#UPHPU [+v jsmith] by anthony.freenode.net |
| 03:18.49 | *** join/#uphpu TodPunk (~todpunk@host-210.pl1071511.fiber.net) |
| 03:18.49 | *** join/#uphpu optikal (optikal@failcore.com) |
| 03:18.49 | *** mode/#UPHPU [+v optikal] by anthony.freenode.net |
| 03:19.15 | *** join/#uphpu raifbot (~raifbot@ec2-184-73-163-211.compute-1.amazonaws.com) |
| 03:19.15 | *** join/#uphpu bradym (~bmitchell@12.91.245.190) |
| 03:19.15 | *** mode/#UPHPU [+v bradym] by anthony.freenode.net |
| 03:19.16 | *** join/#uphpu ninnypants (~ninnypant@69.169.155.194.provo.static.broadweavenetworks.net) |
| 03:19.16 | *** join/#uphpu utahcon (~utahcon@li51-147.members.linode.com) |
| 03:19.16 | *** join/#uphpu romanovic (~romanovic@puritynine.com) |
| 03:19.16 | *** join/#uphpu influx (~influx@deadink.org) |
| 03:19.16 | *** mode/#UPHPU [+v utahcon] by anthony.freenode.net |
| 03:19.35 | *** join/#uphpu dharmatech (~hayduke@166-70-57-177.ip.xmission.com) |
| 03:19.35 | *** join/#uphpu eggyknap (~jtolley@unaffiliated/eggyknap) |
| 03:19.35 | *** mode/#UPHPU [+v eggyknap] by anthony.freenode.net |
| 03:20.10 | *** join/#uphpu mindjuju (~mindjuju@173-10-26-193-BusName-utah.ut.hfc.comcastbusiness.net) |
| 03:20.10 | *** join/#uphpu tierra (~tierra@ipv6.ibaku.net) |
| 03:20.10 | *** join/#uphpu unum (~unum@truecomputing.biz) |
| 03:20.10 | *** mode/#UPHPU [+o mindjuju] by anthony.freenode.net |
| 03:20.33 | *** join/#uphpu carmony (~justin@c-24-10-194-172.hsd1.ut.comcast.net) |
| 03:20.33 | *** join/#uphpu DexterTheDragon (~admin@oalug/member/blackchaos) |
| 03:20.33 | *** join/#uphpu Heartsbane (tsharpe@unaffiliated/heartsbane) |
| 03:20.33 | *** join/#uphpu herlo (~clints@fedora/herlo) |
| 03:20.33 | *** mode/#UPHPU [+vvvv carmony DexterTheDragon Heartsbane herlo] by anthony.freenode.net |
| 03:20.35 | *** join/#uphpu jnbek (~jnbek@pdpc/supporter/active/jnbek) |
| 03:20.35 | *** join/#uphpu synic (~squish@pdpc/supporter/student/synic) |
| 03:20.35 | *** join/#uphpu ags (~ags@mx.vensair.net) |
| 03:20.35 | *** mode/#UPHPU [+v synic] by anthony.freenode.net |
| 03:34.07 | *** join/#uphpu brasto (~brasto@c-67-182-203-21.hsd1.ut.comcast.net) |
| 03:34.15 | *** mode/#UPHPU [+v brasto] by ChanServ |
| 03:52.35 | *** join/#uphpu josephscott (~josephsco@74.200.246.169) |
| 03:52.36 | *** mode/#UPHPU [+v josephscott] by ChanServ |
| 04:14.14 | *** join/#uphpu jtsnow (~jtsnow@VDSL-151-118-130-166.DNVR.QWEST.NET) |
| 04:29.44 | *** join/#uphpu macnewbold1 (~Adium@c-67-166-115-242.hsd1.ut.comcast.net) |
| 04:29.44 | *** mode/#UPHPU [+o macnewbold1] by ChanServ |
| 05:27.16 | josephscott | $chuck-fact |
| 05:27.16 | raifbot | Chuck Norris will never have a heart attack. His heart isn't nearly foolish enough to attack him. |
| 11:46.38 | *** join/#uphpu brasto (~brasto@c-67-182-203-21.hsd1.ut.comcast.net) |
| 11:46.38 | *** mode/#UPHPU [+v brasto] by ChanServ |
| 12:50.41 | *** join/#uphpu jtsnow (~jtsnow@VDSL-151-118-130-166.DNVR.QWEST.NET) |
| 14:00.08 | *** join/#uphpu itimber (~Adium@74.51.216.34) |
| 14:00.08 | *** mode/#UPHPU [+v itimber] by ChanServ |
| 14:06.53 | *** join/#uphpu josephscott (~josephsco@74.200.246.169) |
| 14:06.53 | *** mode/#UPHPU [+v josephscott] by ChanServ |
| 14:17.27 | *** join/#uphpu cro (~cro@204-228-149-217.ip.xmission.com) |
| 14:19.40 | *** join/#uphpu cro (~cro@204-228-149-217.ip.xmission.com) |
| 14:31.25 | *** join/#uphpu redroy (~Adium@67.137.120.26) |
| 14:37.39 | *** join/#uphpu Utah_Dave (~boucha@72-166-51-162.dia.static.qwest.net) |
| 14:37.39 | *** mode/#UPHPU [+o Utah_Dave] by ChanServ |
| 14:43.51 | *** join/#uphpu macnewbold (~Adium@66.60.115.30) |
| 14:43.51 | *** mode/#UPHPU [+o macnewbold] by ChanServ |
| 14:50.11 | *** join/#uphpu Utah_Dave (~boucha@72-166-51-162.dia.static.qwest.net) |
| 14:50.11 | *** mode/#UPHPU [+o Utah_Dave] by ChanServ |
| 14:59.11 | josephscott | $chuck-fact |
| 14:59.11 | raifbot | Chuck Norris is the only man to punch a cyclops between the eyes |
| 15:02.16 | *** join/#uphpu fungus (~olsonl@bromine.sosstaffing.com) |
| 15:02.16 | *** mode/#UPHPU [+o fungus] by ChanServ |
| 15:20.22 | *** join/#uphpu mgeary (~mgeary@166.70.44.49) |
| 15:20.22 | *** mode/#UPHPU [+v mgeary] by ChanServ |
| 15:20.43 | mindjuju | wow, looks like mysql.com hot hacked |
| 15:21.01 | mgeary | mindjuju: ? |
| 15:21.17 | herlo | mindjuju: oh, I think that was the other day... |
| 15:21.35 | mindjuju | yeah, 2 days ago |
| 15:21.50 | mindjuju | i guess it was a busy weekend for the hackers |
| 15:22.01 | mindjuju | @mgeary - http://seclists.org/fulldisclosure/2011/Mar/309 |
| 15:22.25 | josephscott | yep |
| 15:22.34 | mindjuju | poor devils have their database schema posted for the whole world to see |
| 15:23.50 | josephscott | schema is the least of their issues |
| 15:24.36 | DexterTheDragon | who cares about schemas, it's the passwords that are fun |
| 15:24.37 | josephscott | password hashes on the other hand, that is not fun |
| 15:26.37 | *** join/#uphpu TuxToaster (~TuxToaste@173-10-25-150-BusName-utah.ut.hfc.comcastbusiness.net) |
| 15:30.48 | *** join/#uphpu ninnypants (~ninnypant@69.169.155.194.provo.static.broadweavenetworks.net) |
| 15:32.08 | herlo | I don't know, cloud drive doesn't sound that awesome |
| 15:32.11 | herlo | it's nice though |
| 15:33.36 | *** part/#uphpu mgeary (~mgeary@166.70.44.49) |
| 15:47.54 | *** join/#uphpu bhosie (~bhosie@192.84.171.170) |
| 15:51.58 | *** join/#uphpu mgeary (~mgeary@166.70.44.49) |
| 15:51.58 | *** mode/#UPHPU [+v mgeary] by ChanServ |
| 15:52.16 | mgeary | romanovic-afk: ping |
| 16:00.57 | mindjuju | so herlo, when you guys buy computers, where are you buying them from? reseller? cdw? direct? |
| 16:01.25 | influx | Mornin' |
| 16:01.31 | mindjuju | sup influx |
| 16:01.34 | mindjuju | what about you fungus? |
| 16:01.39 | mindjuju | where do you get your comps from? |
| 16:01.53 | herlo | mindjuju: I wasn't involved in that, but we did buy a bunch from hp directly |
| 16:01.57 | stderr- | whoo new X-7 phenom |
| 16:02.03 | herlo | mindjuju: unless you are talking about servers |
| 16:02.10 | mindjuju | nope, desktop |
| 16:02.17 | herlo | we bought from whoever gave the best deal on servers |
| 16:02.40 | herlo | hp blades first time and dell servers second time |
| 16:04.02 | fungus | mindjuju, for personal use, Apple. For company use, Dell direct, or IBM via VAR depending on it's application. |
| 16:04.15 | fungus | oh, desktops. Dell. |
| 16:04.17 | mindjuju | upper management |
| 16:04.23 | mindjuju | that's where i am right now, dell |
| 16:04.52 | herlo | I do like the idea of going with laptops as desktops for the company. It generally makes sense these days |
| 16:05.23 | fungus | That's sounds good in theory |
| 16:05.31 | fungus | but users suck, and break laptops more often. |
| 16:05.32 | mindjuju | until they lose it |
| 16:05.38 | *** join/#uphpu josephscott (~josephsco@74.200.246.169) |
| 16:05.38 | *** mode/#UPHPU [+v josephscott] by ChanServ |
| 16:05.42 | mgeary | yeah, laptops are much more expensive to repair |
| 16:05.57 | mgeary | i always get AppleCare on my laptops. Never get it on desktops |
| 16:06.19 | mgeary | you have a problem with your screen on a (Mac) Laptop, and you're looking at ~$1000 |
| 16:06.23 | *** part/#uphpu jtsnow (~jtsnow@VDSL-151-118-130-166.DNVR.QWEST.NET) |
| 16:06.39 | mindjuju | one thing we did figure out that worked well at BGM is a laptop that VPN + RDP to desktop = happyness and joy |
| 16:06.59 | mindjuju | if they lost the laptop, or stolen or whatever, not really worried about docs lost, etc |
| 16:12.49 | fungus | nice idea, but then you have to buy both, and user experience can be a little complicated. |
| 16:13.14 | fungus | We have lots of bonehead users here. |
| 16:14.16 | herlo | I don't know, it's not like they get to take it home with them and use it other places or anything. The value may outweigh the costs depending on who gets the laptops |
| 16:15.12 | *** join/#uphpu redroy (~Adium@67.137.120.26) |
| 16:34.35 | mgeary | hey, anyone know of an easy way to reclaim disk space after truncating a large InnoDB table? |
| 16:36.39 | josephscott | mgeary: tried optimize? http://dev.mysql.com/doc/refman/5.0/en/optimize-table.html |
| 16:36.54 | herlo | mgeary: rm |
| 16:36.56 | herlo | :) |
| 16:37.03 | mgeary | herlo: lemme try that... |
| 16:37.13 | mgeary | Ah, CRAP |
| 16:37.17 | herlo | lol |
| 16:37.44 | mgeary | josephscott: sure enough. that worked. thanks |
| 16:38.12 | herlo | mgeary: good thing you listened to josephscott :) |
| 16:38.30 | josephscott | np |
| 16:38.47 | josephscott | there are some cases where the only way to shrink an innodb table is to drop and re-create it |
| 16:39.03 | mgeary | yeah, i thought that's what i might be facing |
| 16:39.04 | josephscott | but for cleaning up rows marked as deleted optimize will generally do the trick |
| 16:39.20 | herlo | mgeary: I'm good to listen to on many things, but optimizing mysql is not *yet* one of them :) |
| 16:39.47 | herlo | though, I'm getting better and I know how to run and manage multiple instances of mysqld on a single box |
| 16:40.19 | mgeary | why would you do that? |
| 16:40.32 | herlo | mgeary: development environments |
| 16:40.48 | josephscott | in case you want to read a row more than once ;-) |
| 16:41.17 | mgeary | huh. i've never had a situation where it seemed necessary. Interesting |
| 16:41.59 | herlo | mgeary: it's generally not, but mysql provides a binary to do just that |
| 16:43.44 | josephscott | mysql has gradually gotten better about multiple cores and cpus, so you don't see multiple instances on the same box much any more |
| 16:44.52 | herlo | well, like I said, the value for me is about development environments. Little separate spaces where you can work on code, with a full environment to use. Web server, db server and code. It's beautiful :) |
| 16:45.25 | josephscott | not using VMs for that? |
| 16:45.27 | mgeary | herlo: i just have a local config file that gets included at runtime that can override things like db name, paths to app |
| 16:45.58 | herlo | josephscott: no, that would require more $$ |
| 16:46.07 | josephscott | at work usually each dev has their own VM, to help keep things separate |
| 16:46.25 | herlo | mgeary: but then you have to maintain that, this makes it easy to switch between projects quickly by just changing a directory |
| 16:46.44 | herlo | josephscott: yeah, this is the same idea, but you can have multiple environments on a single box. It works well |
| 16:47.25 | herlo | having a separate vm for each dev per project feature would make a lot of work methinks |
| 16:47.31 | josephscott | do you enforce memory, cpu and disk usage limits, or just no worry about them? |
| 16:48.00 | herlo | well, I've not needed to worry about them yet. I don't think the issue will ever come up tbh |
| 16:48.06 | josephscott | if you had to rebuild each one from scratch, yeah, that would be work, but that is what bash is for ;-) |
| 16:48.13 | herlo | but quotas can be put in place |
| 16:48.27 | herlo | josephscott: bash is limited in what it can do |
| 16:48.46 | josephscott | so I'm curious then, what does have a separate mysqld instance do that using separate DBs in one instance not do for you? |
| 16:48.55 | josephscott | having |
| 16:49.34 | herlo | well, lvm snapshots are much easier to maintain than having to dump a db from live every time I want to create a new dev environment |
| 16:49.54 | herlo | you start up a new db instance and point it at the new snapshot, done |
| 16:50.29 | herlo | if I have to do that in one instance, I have to keep track of the db name, and it's another thing I have to adjust in the code. |
| 16:51.04 | herlo | keeping the code base clean is important, and while you can do it your way, I don't think this way is a bad solution |
| 16:51.16 | josephscott | doesn't your code have to switch to using a different port to talk with mysql? |
| 16:51.19 | herlo | and it's something that's repeatable |
| 16:51.42 | josephscott | I'm just trying to figure out where the wins are for going with multiple instances |
| 16:52.08 | herlo | josephscott: so if you want the whole scenario, it's like this |
| 16:52.21 | herlo | I create a single box for all devs at a company |
| 16:52.25 | herlo | each person has their own login |
| 16:52.44 | herlo | and on demand, the person can create an independent development environment, we call it a camp |
| 16:53.16 | herlo | it includes their own db, their own copy of the code from a repo and all the tools to manage it |
| 16:53.49 | herlo | I think the thing about vms is that while they are nice, unless you've built tools to make them available to a dev on-demand, they would have to setup all of these pieces themselves each time |
| 16:54.08 | herlo | the other nice thing is they can have multiple of these in a single location |
| 16:54.20 | herlo | switch from camp1 to camp4 and you can work on feature X |
| 16:54.56 | herlo | and you can have multiple projects on a single instance too, as long as they have similar requirments, like a webserver and/or database |
| 16:54.57 | josephscott | we basically do the same thing, but geared around VMs |
| 16:55.08 | josephscott | and apply roles to a VM |
| 16:55.11 | herlo | indeed, just a different way to do it, but VMs are costly |
| 16:55.29 | josephscott | so if a dev needs to work on projects a, b, and c then the VM gets roles a, b, and c |
| 16:56.05 | herlo | do you create vhosts for each? how are they separated within the vm? |
| 16:56.08 | josephscott | yep, Xen will add some over head, the trade off is generally a single dev can not bring down the entire system |
| 16:56.37 | josephscott | yes, a dev can map any site to their VM |
| 16:57.06 | herlo | I don't see how a dev can take down the entire system in my way either |
| 16:57.24 | josephscott | haven't filled up /tmp/ recently? |
| 16:57.28 | influx | ~lart Excel |
| 16:57.28 | ibot | keeps mailing Excel free America Online CDs until he drowns |
| 16:57.46 | josephscott | or had a dev create an infinite loop that eats up all the RAM on a system |
| 16:57.59 | herlo | josephscott: but the dev has to manage their own vhosts, no? |
| 16:58.21 | josephscott | no, we don't generally give out root to devs |
| 16:58.31 | herlo | josephscott: all I know is that we did this sort of setup at Backcountry with little to no damage in the way you are suggesting |
| 16:58.48 | herlo | josephscott: so how do they create a new project and manage it on this one vm? |
| 16:58.54 | josephscott | not saying it is guaranteed, just trade offs |
| 16:58.59 | herlo | yeah, for sure |
| 16:59.25 | josephscott | depends on the project, for the most part we don't create tons of new top level projects |
| 16:59.39 | josephscott | they usually come under the scope of one of the existing projects |
| 16:59.45 | herlo | right |
| 16:59.54 | herlo | but you said 10:55 <+josephscott> so if a dev needs to work on projects a, b, and c then the VM gets roles a, b, and c |
| 17:00.04 | herlo | how do roles a, b and c get applied to the vm? |
| 17:00.10 | josephscott | systems team |
| 17:00.40 | josephscott | we are pretty open, but we like to have a request trail for when a dev gets access to a project |
| 17:00.41 | herlo | josephscott: actually, I was looking for the detail of what they did. I mean, do they manually go in and create that stuff for the dev? |
| 17:00.53 | josephscott | it's all scripted |
| 17:01.08 | herlo | josephscott: sure, and camps does too. Integration with bug tracking and the like is in the works |
| 17:01.27 | josephscott | at a high level, it makes sure the right libs are installed, web server configs and checks out source code from the repo |
| 17:01.30 | herlo | the tools are written in a way to make it easy to integrate into a paper trail |
| 17:01.41 | herlo | josephscott: cool, pretty much what camps does |
| 17:02.17 | herlo | I should mention, /tmp doesn't get filled up if you are careful when you set up the initial box. That's part of setup |
| 17:03.01 | josephscott | one thing the VM approach allows as well is that the user account structure is the same for every VM |
| 17:03.09 | herlo | yes |
| 17:03.13 | josephscott | and auth is all done based on ssh keys |
| 17:03.30 | herlo | the camps approach just handles that with individual users setting up ssh keys once |
| 17:03.41 | herlo | and then the project repo determines whether the user has access |
| 17:04.14 | herlo | https://github.com/sitaramc/gitolite/blob/pu/doc/wildcard-repositories.mkd <-- this is the bomb |
| 17:05.09 | herlo | josephscott: so some other things that camps does that are partly convenience features are wrap an scm for them. This was requested by a client |
| 17:06.22 | herlo | but I like making it easy for them to do their job, and not worry about the interface. If they want to use git, they can, but it's nice to have convenience features |
| 17:07.29 | josephscott | they have another layer to abstract source control? |
| 17:07.40 | josephscott | that would be hard core :-) |
| 17:07.56 | herlo | yeah, not that you can't use the scm if you want |
| 17:08.30 | herlo | but it does have some basic wrappers for sharing camps, adding and commiting to the repo and reading the log |
| 17:09.16 | herlo | josephscott: so my project, pycamps (https://github.com/herlo/PyCamps/tree/develop) is based upon a perl project called devcamps (http://www.spreecamps.com/devcamps-tutorial) |
| 17:10.09 | herlo | their project didn't do some of the things pycamps did (scm wrapping, repo sharing, project management) and their commands were all over the place. pycamps has a succinct list of commands and it's easy to figure them out with a bit of trial and error |
| 17:10.48 | herlo | I am just about done with v0.2 of camps, if you have time, try it out. I'd love to have some additional feedback. |
| 17:12.01 | josephscott | at work that sort of thing doesn't work unfortunately, making copies of the DB would be painful and in some cases flat out not allowed (securing private data, etc.) |
| 17:12.40 | herlo | josephscott: yeah, so we actually handle that in an interesting way |
| 17:13.05 | herlo | each night we do a dump from live |
| 17:13.33 | herlo | it gets altered with passwords and the like to match our master camp |
| 17:14.09 | herlo | this is what we snap from. But, if needed, we could easily enough pull another live dump (takes about 45-60 mins) and run it through the same process anytime during the day |
| 17:14.55 | josephscott | how big are your DBs? |
| 17:14.55 | herlo | 10G atm |
| 17:15.16 | herlo | but we're on the same network when we do the dump, the camps are on aws, so it's fast to copy as well |
| 17:15.25 | josephscott | and just one DB for the entire project? |
| 17:15.34 | herlo | well, that's all the dbs |
| 17:15.42 | herlo | there are 4 different projects |
| 17:15.46 | herlo | possibly 5 |
| 17:16.20 | herlo | I suspect the size of the db will be an issue later on |
| 17:16.33 | josephscott | yeah, we have many, many tables that are over 10GB by themselves |
| 17:16.45 | josephscott | and hundreds of DBs |
| 17:16.57 | herlo | yeah, I don't think camps scales to that level yet |
| 17:17.38 | herlo | It is definitely something to consider. Currently, the idea was to remove unneeded data from the db and compact it. |
| 17:18.02 | herlo | josephscott: how do you do db backups? That seems like it could get pretty hairy |
| 17:18.06 | josephscott | I'm still working on fixing some of our DB workflow because MySQL replication can't keep up with the master |
| 17:18.44 | josephscott | different backup methods/schedules depending on the DB/usage. file system level snapshots for many |
| 17:19.15 | herlo | thought so |
| 17:19.24 | herlo | replication is a bit pita |
| 17:19.31 | herlo | but it's nice when it's working properly |
| 17:19.58 | josephscott | we use it quite a bit, but yes, is a pain when the master is doing so many inserts/updates that the slave can never catch up |
| 17:20.01 | herlo | do you not have fiber channel between your db servers to do replication? |
| 17:20.06 | jsmith | And when it breaks -- it breaks spectacularly! |
| 17:20.12 | herlo | yes it does! |
| 17:20.38 | herlo | josephscott: not saying that fiber will solve your problems, but iirc you guys don't own your hardware? |
| 17:21.09 | josephscott | herlo: have you tried mysql replication across different DCs in different states? |
| 17:21.16 | herlo | no |
| 17:21.19 | herlo | scares me |
| 17:21.28 | josephscott | that would be one really long cable :-) |
| 17:21.35 | herlo | josephscott: which is why I asked |
| 17:21.46 | josephscott | we have dedicated hardware in each DC |
| 17:21.50 | josephscott | and multiple DCs |
| 17:22.22 | herlo | right, so replication between is the pain point |
| 17:22.44 | josephscott | one server that is having this problem averages around 2.4k updates and 350 inserts per second |
| 17:22.52 | herlo | nice |
| 17:22.55 | mindjuju | wow |
| 17:23.12 | josephscott | the master keeps up fine, but the slave never catches up |
| 17:23.27 | josephscott | single threaded replication, only so much you can do |
| 17:23.39 | herlo | right |
| 17:23.41 | josephscott | so making some adjustments to compensate a bit |
| 17:24.08 | herlo | what about localized master/master setup? |
| 17:24.39 | herlo | then only the writes to one master get sent to the other master (as a slave) |
| 17:24.41 | josephscott | master master has a separate set of issues |
| 17:24.50 | herlo | so you've tried that too? |
| 17:25.07 | josephscott | yeah, for some conditions master master in mysqld can be much worse |
| 17:25.17 | josephscott | and the multi DC situation doesn't make that any easier |
| 17:25.26 | herlo | indeed |
| 17:25.39 | herlo | just read about them a bunch |
| 17:25.43 | DexterTheDragon | master/master is muuuch worse |
| 17:25.59 | herlo | we did master / slave only |
| 17:26.26 | josephscott | have I mentioned that I really like munin lately |
| 17:26.40 | josephscott | love having graphs |
| 17:26.44 | herlo | munin rox |
| 17:26.58 | josephscott | DexterTheDragon: hehe, tried master master did you? |
| 17:27.18 | DexterTheDragon | we used to use master master here at work |
| 17:28.16 | josephscott | did it blow up nicely? |
| 17:28.46 | DexterTheDragon | that thing would get out of sync like once a month and when it was out of sync you were boned |
| 17:28.52 | DexterTheDragon | duplicate records all over the place |
| 17:29.17 | mindjuju | sounds painful |
| 17:29.42 | josephscott | yep |
| 17:30.00 | josephscott | master master sounds more attractive on paper than it usually ends up being in practice |
| 17:30.41 | DexterTheDragon | and when a user double clicked a form submit it could do this cool thing where each request would connect to a different master |
| 17:31.11 | josephscott | awesome |
| 17:31.29 | mindjuju | :) |
| 17:31.44 | DexterTheDragon | thats usually how the out of sync started |
| 17:32.10 | DexterTheDragon | replicator would hit a unique key restraint then decide to die |
| 17:34.04 | josephscott | oh, and real time notification in irc via bots, I love that too |
| 17:34.21 | josephscott | realtime ish at least |
| 17:40.10 | josephscott | I also really like it when I deploy something new and it doesn't blow up in my face :-) |
| 17:47.27 | mindjuju | does anybody have a favorite nagios web interface i should look at? |
| 17:52.03 | fungus | I just use the ugly stock interface with the pnp addon |
| 17:53.30 | TuxToaster | $rlart |
| 17:53.31 | raifbot | urinates on cedwards. |
| 17:53.41 | TuxToaster | raifbot has no manners |
| 17:54.03 | TuxToaster | apparently neither does fungus |
| 17:54.22 | TuxToaster | :P |
| 17:54.32 | fungus | nope. :) |
| 18:02.11 | *** join/#uphpu jkr801 (~jkr801@dsl093-228-008.slc1.dsl.speakeasy.net) |
| 18:05.52 | influx | How close is the next UPHPU meeting? |
| 18:07.20 | herlo | influx: always on third thursday |
| 18:08.09 | Utah_Dave | hey what was that command line program beandog used in his presentation to show a progress bar? |
| 18:09.10 | Utah_Dave | was it pv? |
| 18:11.00 | fungus | yes, pv |
| 18:12.34 | Utah_Dave | ok, thanks! |
| 18:20.10 | *** join/#uphpu dharmatech (~hayduke@173-165-135-245-utah.hfc.comcastbusiness.net) |
| 18:20.44 | mindjuju | utah_dave, you around? |
| 18:32.40 | mindjuju | fungus, you still around? |
| 18:32.58 | fungus | yup |
| 18:33.00 | mindjuju | if you had things to do over again, would you do nagios or zenoss ? |
| 18:33.32 | fungus | I haven't fully investigated zenoss yet. But i would definitely give it a try. |
| 18:34.07 | fungus | However, IIRC Nagios had much better flexibility, and in my environment, that is very important. |
| 18:34.35 | mindjuju | rebuild nagios sound easier as the server components are already in place |
| 18:35.04 | Utah_Dave | mindjuju: hey, I'm back |
| 18:35.20 | fungus | Since you got most of the learning curve of nagios done, that makes sense. The learning curve is the biggest obstacle to nagios IMHO |
| 18:35.32 | mindjuju | ah utah_dave, i'm in a conundrum |
| 18:35.56 | mindjuju | our voip provider / support company has informed me that after much research, they can't offer me a VPN phone like they promised |
| 18:36.03 | mindjuju | i'm pissed! grrr! |
| 18:36.06 | Utah_Dave | fungus: yeah, nagios has been a headache to get all set up for us. We've got it mostly working now, but it seems it shouldn't be so hard to get going |
| 18:36.28 | mindjuju | do you know of a VPN phone that i should look into |
| 18:36.46 | mindjuju | i know they are mainly based on the phone server software, so i need one that is flexible |
| 18:36.56 | Utah_Dave | do you mean a voip phone that will connect to a vpn first and then connect to it's sip provider? |
| 18:36.57 | fungus | Yup, and the initial config of nagios, if done right, makes general maintenance really easy. |
| 18:37.10 | mindjuju | yup, exactly utah_dave |
| 18:37.54 | mindjuju | i started to google, but thought i'd get a recommendation |
| 18:37.56 | Utah_Dave | I haven't used one yet. |
| 18:39.13 | mindjuju | so utah_dave, what do you think about VOIP over VPN, is goot? |
| 18:39.31 | mindjuju | guy is in florida, server here with me in utah |
| 18:39.48 | Utah_Dave | I've heard it can solve some audio issues because vpn is often given qos priority in upstream routers |
| 18:40.15 | Utah_Dave | You can get a small router to do the vpn and then get a regular voip phone behind it |
| 18:40.33 | mindjuju | i was just starting to wonder about that |
| 18:40.42 | mindjuju | some kind of router / vpn |
| 18:42.36 | mindjuju | @utah_dave, do you have a router that you recommend? else, i'll look at linksys, theys my favorite |
| 18:44.29 | Utah_Dave | I'd get a linksys and put dd-wrt on it. |
| 18:45.09 | Utah_Dave | mindjuju: I have to run to a meeting, but I'll be back on in a couple hours. |
| 18:45.17 | mindjuju | thx! |
| 18:45.37 | Utah_Dave | If you get a separate router, then you can get whatever voip phone you want and not have to worry about vpn capabilities in the phone |
| 18:46.03 | Utah_Dave | that's what I'd do, unless you REALLY need the phone to handle all of it. I've never bothered |
| 19:24.29 | *** join/#uphpu macnewbold (~Adium@66.60.115.30) |
| 19:24.29 | *** mode/#UPHPU [+o macnewbold] by ChanServ |
| 19:24.37 | macnewbold | mindjuju: you may consider having him use a voip phone there and talking to your server/pbx over IP, rather than sending the traffic through VPN - the QoS prioritization would likely get nullified if it was VOIP wrapped in VPN rather than just plain VOIP packets. |
| 19:28.45 | *** join/#uphpu brasto (~brasto@c-67-182-203-21.hsd1.ut.comcast.net) |
| 19:28.45 | *** mode/#UPHPU [+v brasto] by ChanServ |
| 19:29.02 | fungus | QoS is usually nullified in transport by upstream providers. |
| 19:29.32 | fungus | Also the 3-letter organizations can listen to your calls. ;) |
| 19:31.20 | fungus | And for an awesome firewall/VPN/router. Cisco ASA 5505. |
| 19:51.45 | *** join/#uphpu redroy (~Adium@67.137.120.26) |
| 20:06.50 | *** join/#uphpu redroy1 (~Adium@67.137.120.26) |
| 20:13.10 | influx | Oi. |
| 20:13.45 | influx | I've put together a script that exports the results of a database query to csv and magically presents the user with a download prompt. |
| 20:14.08 | mgeary | ~admire influx |
| 20:14.09 | ibot | ACTION secretly and androgynously wishes it were just like influx |
| 20:14.26 | influx | But, I've run into a teensy snag. |
| 20:15.38 | influx | Right now, the file is downloaded via anchored link "Download results as CSV". If the user clicks it again, the browser will go the page that contains the script and I'll see errors about header files already having been sent. |
| 20:16.13 | influx | http://pastebin.com/Qg53VpcH |
| 20:17.10 | influx | I'm wondering about what the problem might be, and how to solve it. |
| 20:17.52 | influx | I'm pretty sure it's possible to reset things without having to leave the results page, but I'm unsure of where to begin. |
| 20:18.06 | influx | Any thoughts, chaps? |
| 20:27.22 | unum | influx: what's happing is that your errors are being sent before your header commands |
| 20:28.09 | unum | the reason you get errors is because you delete the session variable 'csv_to_download' |
| 20:28.16 | unum | what is the behavior you want? |
| 20:30.31 | influx | To allow the user to download the CSV file again if they haven't left the page that links to it. |
| 20:30.43 | influx | But clear the session variable when they're done. |
| 20:31.07 | unum | then you should clear the variable with an ajax from onunload |
| 20:31.09 | influx | So, I guess, I'll have to figure out where a good place for the session-clearing script should reside. |
| 20:31.49 | influx | from or form? |
| 20:31.56 | influx | *curious* |
| 20:32.03 | unum | though I'd probably clear it in the code to the page that lists the the possible files to download |
| 20:32.45 | unum | the option I mentioned earlier: <body onunload='clear()'> |
| 20:32.55 | unum | but I wouldn't use that one personally |
| 20:33.44 | unum | I would delete the variable on the "menu" page |
| 20:34.05 | unum | of couse I would allow them to just download from the menu page myself and not force them to a seperate page |
| 20:34.35 | unum | unless you are of course doing it provide information, advertising, or as a learning tool |
| 20:37.18 | unum | so what is your reason for directing them to a seperate page before they download? |
| 20:42.39 | influx | One sec. |
| 20:43.16 | influx | They're on a page where they can see their sales numbers between a date range that they specify in a form. |
| 20:43.57 | influx | The form being on the same page. By default, it shows the numbers for all of the month so far. |
| 20:44.33 | influx | So, if they need a spreadsheet of the data, they can click a link and it should prompt them to download the exported file, without having them leave the page. |
| 20:47.30 | *** join/#uphpu redroy (~Adium@67.137.120.26) |
| 20:53.37 | influx | There's only ever the one file to download--the one that contains the most recent mysql query results |
| 21:03.07 | mindjuju | ~give influx pat on the back |
| 21:03.07 | ibot | ACTION gives pat on the back to influx |
| 21:03.10 | influx | Going to reset a flag at the head of the file |
| 21:03.30 | influx | depending on whether or not the flag has been set. |
| 21:03.39 | influx | then I'm going to set the flag to "on" let's say at the end of the file. |
| 21:04.20 | influx | If the flag is set to on by next page load, then the session variable from earlier will be cleared, allowing for a new one to be set. |
| 21:05.11 | influx | Though I guess that still leaves the session variable populated if they leave the page after getting what they need. |
| 21:05.31 | influx | mindjuju: Thanks. |
| 21:06.56 | influx | ~cookie mindjuju |
| 21:07.46 | influx | Didn't figure you for an equines enthusiast |
| 21:08.27 | mindjuju | i'm native Texan. all native Texans have a backup car incase horse goes down |
| 21:14.09 | josephscott | haha! http://devour.com/video/the-ultimate-warrior-meets-arsenio-hall/ |
| 21:19.06 | influx | Heh. |
| 21:23.27 | *** join/#uphpu utahcon_opera (~whatisthi@205.204.33.82) |
| 21:23.47 | *** part/#uphpu utahcon_opera (~whatisthi@205.204.33.82) |
| 21:24.02 | *** join/#uphpu utahcon_opera (~whatisthi@205.204.33.82) |
| 21:24.17 | utahcon | lol -- nope that won't get annoying :| |
| 21:24.22 | *** part/#uphpu utahcon_opera (~whatisthi@205.204.33.82) |
| 21:27.54 | unum | influx: if there is only ever one file why do you have the variable? |
| 21:28.15 | unum | oh I think I understand now |
| 21:28.57 | unum | personally though I'd handle it with a get request |
| 21:37.39 | influx | The variable contains an array returned by a function. |
| 21:42.57 | mindjuju | josephscott, never pegged you as a WWF watcher |
| 21:43.21 | josephscott | I'm not, but watching that old clip was funny |
| 21:43.26 | mindjuju | ah |
| 21:43.33 | josephscott | with arsenio trying to stay out of his way |
| 21:43.37 | mindjuju | yup |
| 21:43.51 | mgeary | i totally would watch WWF to wind down after coming home from closing up at Taco Bell at 1:30am |
| 21:44.04 | mgeary | sigh. those were days |
| 21:44.10 | mindjuju | really? |
| 21:44.36 | mindjuju | i forget that you served time at tbell |
| 21:44.59 | josephscott | haha, "served time" |
| 21:45.41 | josephscott | well, I worked at McD's some during high school, paid the few bills I had |
| 21:46.39 | mgeary | yeesh |
| 21:47.41 | mindjuju | i hear they make beans from flakes now (like mash potato flakes) |
| 21:48.53 | mgeary | that's how it works at the LDS cannery as well |
| 21:49.00 | mgeary | i have some of those in my food storage |
| 21:52.23 | *** join/#uphpu dharmatech (~hayduke@166-70-57-177.ip.xmission.com) |
| 21:52.55 | mindjuju | so josephscott, can you tell us what the "secret sauce" is for the big mac? |
| 21:53.02 | mindjuju | i heard it was thousand island dressign |
| 21:53.44 | *** join/#uphpu macnewbold1 (~Adium@66.60.115.30) |
| 21:53.44 | *** mode/#UPHPU [+o macnewbold1] by ChanServ |
| 21:58.25 | *** join/#uphpu romanovic (~romanovic@puritynine.com) |
| 22:03.50 | carmony | watched a video about node.js |
| 22:03.53 | carmony | pretty cool stuff |
| 22:06.07 | josephscott | very |
| 22:06.34 | josephscott | carmony: might finally get around to testing redis for potential production use next month |
| 22:06.39 | carmony | oh yeah? |
| 22:06.42 | influx | hmm |
| 22:07.03 | josephscott | going to see how it compares to the 12GB innodb table I'm dealing with now :-) |
| 22:08.09 | influx | I could use cookies with a time limit. |
| 23:10.28 | *** join/#uphpu brasto (~brasto@70-0-178-20.pools.spcsdns.net) |
| 23:10.28 | *** mode/#UPHPU [+v brasto] by ChanServ |
| 23:14.48 | *** part/#uphpu mgeary (~mgeary@166.70.44.49) |
| 23:17.09 | *** join/#uphpu brasto1 (~brasto@174-149-17-129.pools.spcsdns.net) |
| 23:17.09 | *** mode/#UPHPU [+v brasto1] by ChanServ |
| 23:21.52 | mindjuju | herlo, you around? |
| 23:23.06 | herlo | hi |
| 23:23.13 | herlo | mindjuju: yes, I'm here |
| 23:23.17 | mindjuju | hey, what are you using for IT asset management? |
| 23:23.33 | herlo | uh |
| 23:23.42 | mindjuju | and does it allow you to print out some kind of identification labels? |
| 23:23.45 | herlo | I think we used an excel spreadsheet |
| 23:23.49 | mindjuju | cause that's what i want |
| 23:23.52 | mindjuju | hehe |
| 23:23.54 | mindjuju | :) |
| 23:24.05 | herlo | but we did also use rackmonkey |
| 23:24.23 | mindjuju | i was using rackmonkey, but it never got fully implemented |
| 23:24.29 | fungus | mindjuju, don't focus on looking for a solution w/ label printing, that's unnecessary |
| 23:24.40 | mindjuju | yeah? |
| 23:24.48 | fungus | Just buy some asset tag labels, and put that number into your asset tracking system |
| 23:25.02 | mindjuju | oh gotcha |
| 23:25.07 | mindjuju | brb |
| 23:25.24 | fungus | Otherwise you'll get stuck w/ a crappy system because it offers a minor feature. |
| 23:25.33 | herlo | what fungus said |
| 23:26.04 | herlo | rackmonkey is nice though, you can put asset tags in there since they have some customizable fields |
| 23:26.20 | josephscott | don't kid yourself, you'll probably get stuck with a crappy system either way ;-) |
| 23:26.51 | herlo | haha |
| 23:26.53 | fungus | Our accounting department uses a custom access db. |
| 23:27.14 | josephscott | $8ball are there non-crappy asset systems? |
| 23:27.15 | raifbot | josephscott - Without a doubt. |
| 23:27.23 | josephscott | $8ball can you recommend one? |
| 23:27.23 | raifbot | josephscott - My reply is no. |
| 23:27.23 | fungus | as an example of how simple options can achieve your goals |
| 23:27.40 | herlo | fungus: access is simple? |
| 23:27.49 | josephscott | on that note, time for dinner |
| 23:27.51 | fungus | For accounting type guys, yeah |
| 23:27.51 | herlo | simplistic, yes. simple, no. |
| 23:27.56 | herlo | :) |
| 23:28.15 | fungus | accounting people *love* their Excel and Access crap |
| 23:30.01 | herlo | yes, they do |
| 23:32.29 | mindjuju | k, so just get some asset labels and mesh them into rackmoney |
| 23:32.32 | mindjuju | err, monkey |
| 23:32.33 | mindjuju | can do |
| 23:35.46 | TuxToaster | $rlart |
| 23:35.46 | raifbot | slams macnewbold1 against a large cement Tux. |
| 23:35.54 | TuxToaster | ouch |
| 23:36.19 | herlo | rackmoney sounds better though :) |
| 23:36.19 | TuxToaster | and for the earlier mishap with manners... |
| 23:36.25 | TuxToaster | $lart fungus |
| 23:36.25 | raifbot | crushes fungus with a full height scsi disk. |
| 23:36.26 | TuxToaster | :P |
| 23:36.42 | TuxToaster | doh |
| 23:36.52 | TuxToaster | ~shun fungus |
| 23:36.52 | ibot | You are dead to me, fungus. |
| 23:37.02 | TuxToaster | lol |
| 23:37.27 | TuxToaster | put away the garden hose. :P |
| 23:38.10 | herlo | gross! |
| 23:38.41 | TuxToaster | $lart fungus |
| 23:38.42 | raifbot | takes out a seltzer bottle and sprays fungus in the face. You know, one of those old-school seltzer bottles clowns have? Yeah those. Anyway, consider yourself spritzed. |
| 23:38.56 | TuxToaster | haha. ironic lart. |