Activity Graph

Page 1 of 1


[18:21:58] warreng: has left #ruby: ()


[21:30:50] warreng: has joined #ruby
[21:30:50] warreng: Changing host


[05:27:26] warreng: Ping timeout: 252 seconds
[05:31:36] warreng: has joined #ruby


[02:34:42] warreng: Ping timeout: 272 seconds
[02:35:22] warreng: has joined #ruby
[21:09:59] warreng: Ping timeout: 246 seconds
[21:12:34] warreng: has joined #ruby


[15:22:37] warreng: Ping timeout: 255 seconds
[15:23:24] warreng: has joined #ruby


[05:23:26] warreng: Ping timeout: 252 seconds
[05:23:29] warreng: has joined #ruby
[05:23:29] warreng: Changing host


[22:51:05] warreng: i'm going to pastebin a link to my gist, hold on a sec
[22:53:10] warreng: tin123: it has a newline in it??? do "puts personinput.inspect"


[20:05:30] warreng: has joined #ruby
[20:05:30] warreng: Changing host


[00:26:49] warreng: Ping timeout: 256 seconds
[00:30:53] warreng: has joined #ruby
[22:11:33] warreng: Ping timeout: 256 seconds
[22:12:25] warreng: has joined #ruby


[22:21:00] warreng: i need help! i have async workers in a production app that 12 hours ago started eating memory like crazy??? no code changed, but all of a sudden these instances go from 200mb to 1gb in about 20 seconds??? what's the quickest way i can get some dumps on memory usage and see where that 800mb is going?
[22:21:51] warreng: this isn't a "slow leak overtime", it's a "holy crap, something is leaking like crazy and the process crashes in a few minutes"
[22:22:10] warreng: gizmore: haha, thanks
[22:22:21] warreng: zenspider: yes, everytime i start up the process, it grows to 1gb in < 30 seconds
[22:22:48] warreng: (there's a lot of jobs queued up via delayed jobs, and it starts processing them and then runs out of memory in 30 seconds)
[22:23:19] warreng: gizmore: link?


[00:04:05] warreng: ActionView::Template::Error: Couldn't load the Unicode tables for UTF8Handler (failed to allocate memory), ActiveSupport::Multibyte is unusable
[00:04:15] warreng: i just got that on a production server... retried and it went away
[00:04:23] warreng: has anybody seen anything like that before?
[00:04:34] warreng: (i was using a console when it happened!)


[01:18:56] warreng: so is multi_json not affected by the json stuff announced today?


[00:49:02] warreng: autoloading question... i have a rails model called "class Registry" and i have a second model that inherits "class Foo < Registry".... this works fine in dev/prod mode but when i try to run a test with rspec, it ends up using RSpec::Core::SharedExampleGroup::Registry rather than my own Registry model.... how do I fix that?
[00:50:20] warreng: Radar: tried that... didn't help
[00:51:06] warreng: Radar: using :: *should* but it was still using the other one
[00:51:07] warreng: no idea why


[17:20:42] warreng: i'm all patched up now but still am curious how remote-code-execution could happen... couldn't get it to work through a Proc object... has a working example been released yet?
[17:24:46] warreng: gotcha.. i figured as much... was just curious if it'd been announced yet
[17:32:45] warreng: lejonet: actually, i heard it was discovered last week... so we're in day 4 or 5 or something already
[17:37:55] warreng: lejonet: except this isn't "off by default" :-D
[17:38:58] warreng: lejonet: my understanding is every rails app since 2.0+ is vulnerable by default, even if you don't have a single controller in the app