Deathray | Hi does someone here have a couple of minutes to assist me with a problem I've been having lately with my LAMP server in the EC2 cloud? | 00:22 |
---|---|---|
Deathray | It can or can not be simple, but since I am a Linux amateur I do not know how to start. | 00:23 |
Amivit | Sorry i'm deathray, but got disconnected if someone answered me. | 00:27 |
warzauwynn | Amivit: best to just state your problem, not ask for help. that way somebody can answer if they know the answer. | 00:35 |
Amivit | Ok, well, I have an ubuntu-server instance running in the Amazon ec2 cloud, functioning as a LAMP with very default setup (used tasksel and made very few changes). Once or twice a week the server rejects all connections through everything, including ssh and the actual webserver. I can see in the AWS control panel that the CPU is topping at 100%, and my fix is to force a reboot on the instance. But i'm a bit unsure how to f | 00:37 |
Amivit | ind the cause since I can't even SSH into the instance when it's in that state, which is why I was hoping for some guidance here :) | 00:37 |
Amivit | Worst part is that i monitor it with cloudkick which sends me a text whenever it goes offline and recovers again, so sometimes I can wake up and my phone is completely crashed because the inbox is filled with thousands of these texts :P | 00:45 |
adam_g | Amivit: what version of ubuntu? and which kernel version? | 00:59 |
jmarsden | Amivit: Unless you are running thousands of these instances (in which case you are not a beginner!), why do you get thousands of messages per outage? That suggests you have misconfigured the monitoring tool. You should get ONE message when an instance becomes unresponsive, and ONE when it comes back. | 01:03 |
Amivit | jmarsden: Yes I do, but I guess since the cpu is stuck at 100% it occasionaly responds which is why I receive so many messages, anyway that is a minor issue which I obviously fixed my changing the way cloudkick notifies me. adam_g let me check :) | 01:05 |
jmarsden | <Amivit> Worst part is ... now you say "that is a minor issue" ... make your mind up :) I was responding about the thing you said was the worst part... | 01:06 |
Amivit | Ah yeah it was more of a joke since I found it quite funny my old LG crashed because of it.. Im more concerned about my viewers not able to see my blog :) | 01:07 |
jmarsden | Amivit: OK. if you keep an ssh session open to the server running top -d 10 you would perhaps see what was using the CPU when it became unresponsive? Or you can use sar from the sysystat package to get some statistics on CPU use, etc. | 01:11 |
tarzxvf | Amivit, or install something like munin to get a global overview of your system during this events | 01:13 |
jmarsden | Or a cron job running evern 5 minutes that does ps axuww >>/var/log/mylog.txt or similar, so you can read that log later and see what processes were using all the CPU... | 01:21 |
jmarsden | Lots of possibilities :) | 01:21 |
mrroth_ | Is thier any type of premade appliance database /web frontend program were you can keep a record of how you went (the steps) to fix a problem (trouble,shoot) (like observe problem firsthand) and (document each step). | 01:39 |
jmarsden | mrroth_: In a sense, every bug tracker or issue tracker app does that. You could use every blog app to do it too, if you wanted... | 01:41 |
mrroth_ | oh I wanted something for sysadmin, but yea a blog app may do | 01:43 |
mrroth_ | I have trouble explaning what I did from step a to z | 01:44 |
mrroth_ | and if the problem ever happens againI waste time, becuase I never documented, or no one ever did if they had suchproblem. | 01:44 |
tarzxvf | mrroth_ , i think a wiki could handle this | 01:50 |
jmarsden | For a sysadmin, a set of text files can handle this :) :) | 01:51 |
tarzxvf | Lots of possibilities :) | 01:51 |
mrroth_ | thanks | 01:55 |
mrroth_ | yea | 01:55 |
mrroth_ | just email me what I did | 01:55 |
mrroth_ | or make my tickets more clear | 01:56 |
mrroth_ | I useally use the ticket system only for billing purposes | 01:56 |
Jasonn | I have an install error when I try to install openvps-as on my server, http://pastebin.ubuntu.com/632102/ | 03:18 |
jeeves | why is my bind server giving me this? "error (unexpected RCODE REFUSED) resolving " | 03:35 |
=== jeeves is now known as jeeves_moss | ||
=== Jasonn is now known as JasonnAWAY | ||
=== JasonnAWAY is now known as Jasonn | ||
jeeves_moss | how can I setup SSL on a MySQL server on non native networks? | 04:46 |
=== guampa1 is now known as guampa | ||
MACscr | i just upgraded an ubuntu 8.04 lts xen guest to 10.4 lts and rebooted. This error mean anything to any of ya? http://pastebin.com/4SCYFHg1 | 07:31 |
MACscr | i think it might be the migration of grub to grub2, but im not 100% sure how to fix it since i cant use a live cd on it. I can mount the file system on the xen host, but its a centos system, so i cant run grub-pc and grub-common on it | 07:32 |
MACscr | my theory could be completely wrong as well | 07:32 |
jmarsden | Ok, so you have: ERROR Invalid kernel: xc_dom_probe_bzimage_kernel: kernel is not a bzImage | 07:35 |
jmarsden | So... is that file really a bzImage kernel? What does file /boot/vmlinuz-2.6.24-28-xen say? | 07:35 |
MACscr | vmlinuz-2.6.24-28-xen: gzip compressed data, from Unix, last modified: Fri Feb 11 13:32:43 2011, max compression | 07:36 |
jmarsden | zgip compressed, not bzImage. So there you go. | 07:37 |
MACscr | right, but where the heck is it specifying a bzimage | 07:37 |
jmarsden | The stock Ubuntu kernels are bzImages... I don't know who told it to expect one... | 07:37 |
MACscr | im confused. These would have just be installed by apt during the upgrade | 07:38 |
jmarsden | So maybe the normal Ubuntu kernels are bzImage files and the -xen ones are not?? | 07:39 |
MACscr | i have another system running 10.4, but they are 32 bit and running pae kernels | 07:40 |
jmarsden | I'm not familiar with Xen, I'm just trying to work backwards from the error message you saw. | 07:40 |
jmarsden | If you run file /boot/whatever on those, are those kernels bzImage files? | 07:40 |
jmarsden | I'm running a 10.04 amd64 desktop here and it definitely does have a bzImage kernel: /boot/vmlinuz-2.6.32-32-generic: Linux kernel x86 boot executable bzImage, version 2.6.32-32-generic (buildd@allsp, RO-rootFS, root_dev 0x6801, swap_dev 0x3, Normal VGA | 07:41 |
MACscr | vmlinuz-2.6.32-22-generic-pae: Linux kernel x86 boot executable bzImage, version 2.6.32-22-generic-pae (buildd@p, RO-rootFS, root_dev 0x6801, swap_dev 0x3, Normal VGA | 07:41 |
jmarsden | Yup. bzImage is there. | 07:41 |
MACscr | hmm, so how should i go about getting this new ones? what all files do i need to get for the new kernel? | 07:41 |
jmarsden | So the question becomes, why did your upgrade somehow (apparently!) install you a non-bzImage kernel? Does xen do anything funny and run guest kernels from somewhere outside the guest filesystem? | 07:42 |
jmarsden | I think your issue may be in some way xen-related, and I know almost nothing about Xen, unfortunately. | 07:42 |
MACscr | jmarsden: im running a setup (its common) that allows guests to run their own kernels, so no idea why it flaked like this. | 07:42 |
jmarsden | Um. Wait. You are booting the old 8.04.3 kernel | 07:44 |
jmarsden | Line 28 of your pastebin: Booting 'Ubuntu 8.04.3 LTS, kernel 2.6.24-28-xen' | 07:45 |
MACscr | i dont have a newer xen kernel though in my /boot | 07:46 |
MACscr | hmm | 07:46 |
MACscr | very odd | 07:46 |
jmarsden | Lucid kernels are 2.6.32 or so... | 07:46 |
MACscr | yeah, i have the xen-server kernels in there, but not just -xen | 07:46 |
jmarsden | Is it possible Lucid dropped xen support somehow, and you need to create your own -xen kernel for it?? | 07:47 |
jmarsden | MACscr: Is https://help.ubuntu.com/community/Xen#Ubuntu%27s%20Support%20of%20Xen relevant? | 07:49 |
jmarsden | I need to go to bed, but hopefully something I've said will turn out to be useful :) | 07:50 |
MACscr | jmarsden: not relevant. Thanks though. | 07:51 |
mncl-node | good night all :) | 07:51 |
mncl-node | i have an iptable/ perhaps proxy question | 07:55 |
=== Jasonn is now known as JasonnAWAY | ||
=== twister004_ is now known as twister004 | ||
=== twister004_ is now known as twister004 | ||
Everglades | hi | 10:45 |
Everglades | does anybody know how to install apache + mysql + php >_<? | 10:45 |
dutchie11 | during intital install it will ask if you want a web server otherwise you will need to use sudo apt-get install | 10:48 |
Everglades | not really | 10:49 |
Everglades | :/ | 10:49 |
Syria | Hello! I have Ubuntu 10.04 installed on my vps, but the server is not sending any emails! how can i fix this pelase? | 13:22 |
julian | install an mta? | 13:26 |
julian | .o(or start paying a real admin to do stuff for you) | 13:27 |
Syria | o.0 | 13:28 |
Syria | I am a real admin but i am not a professional one. :( | 13:28 |
Syria | ? | 13:31 |
luite | probably read some postfix howto on the community documentation site :) | 13:32 |
luite | email is imho one of the more tricky services to get right unless you have really basic needs | 13:33 |
Syria | luite: Thank you, I will read about that thing called postfix. | 13:33 |
Syria | Scary!! | 13:33 |
Syria | luite: Installing postfix on my vpx solved the problem all I had to do is choosing "Internet Site" thank you very much. :D | 13:40 |
luite | are you using your isp's (or gmail or something) servers to receive mail? | 13:41 |
shauno | Syria: I'd really recommend working thru the documentation to understand what you've just done. I have a similar setup, and it's hit by people trying to use it to relay spam on a near-constant basis | 13:41 |
Syria | shauno: I need it because i have a wordpress blog and it is not sending emails or notifications. | 13:42 |
shauno | I'm not saying don't do it; just take a few minutes more to make sure it's not available as an open relay to spammers | 13:43 |
Syria | shauno: thank you, I will take this into consideration i did not know about that. | 13:44 |
Syria | I don't want to help spammers. :( I hate what they do. | 13:44 |
shauno | that sounded grumpier than it needed to, but I think 'internet site' is a full inbound+outbound setup :/ | 13:47 |
=== shauno_ is now known as shauno | ||
RoyK | anyone that knows a tool to extract i-frames from a video? | 14:39 |
qman__ | shauno, the default postfix "internet site" config allows relay from localnets only | 16:03 |
qman__ | which is pretty much exactly what you want in your own network, but on a remote host, could be an issue | 16:04 |
shauno | sounds like a sane start, but still overkill for a machine that only wants php to beable to mail outbound | 16:05 |
_torn | hi, when I apt-get install php5 it also grabs and tries to start up apache | 16:43 |
_torn | i'm already using nginx, how can i delete/ stop this from happening? | 16:43 |
SpamapS | _torn: php5 pulls in libapache2-mod-php5 by default | 16:49 |
SpamapS | _torn: if you want to use it with nginx, install php5-fpm | 16:49 |
SpamapS | _torn: the 'php5' package is just a "metapackage" | 16:49 |
nobitanobi | Hello. I've installed lamp-server in Ubuntu 11.04, but when trying to access to a php file, instead of that the browser gives me the php file to download.. | 18:13 |
=== whaley is now known as waz_ | ||
=== waz_ is now known as whaley | ||
jmarsden | nobitanobi: And the file is under /var/www/ somewhere? | 18:27 |
nobitanobi | jmarsden: apparently the mod_php was not well installed. | 18:27 |
nobitanobi | thanks | 18:27 |
jmarsden | nobitanobi: You're welcome :) | 18:27 |
nobitanobi | jmarsden: is it a good strategy to create symlinks from /var/www to a folder where I have write permissions? | 18:28 |
nobitanobi | so I don't have to keep writting in /var/www as root | 18:28 |
jmarsden | nobitanobi: You can create sites as needed underneath /var/www that have the permissions you need, say /var/www/nobitanobi/ and configure them as sites for your apache to serve up. | 18:30 |
nobitanobi | ok | 18:31 |
jmarsden | Symlinks.... well, it should work, but it's one more layer of indirection for every page apache serves up... | 18:31 |
qman__ | doing it with symlinks is risky, you'd be better off creating additional websites in the user's home | 18:31 |
nobitanobi | I will do the jmarsden solution then | 18:33 |
nobitanobi | creating those folders and configuring them | 18:34 |
qman__ | good plan | 18:34 |
nobitanobi | thanks guys | 18:34 |
nobitanobi | the idea is creating a folder where my user can write | 18:35 |
nobitanobi | right? | 18:35 |
nobitanobi | but to make apache able to serve the contents, what is needed to do? | 18:35 |
qman__ | www-data needs read and possibly execute permission | 18:36 |
qman__ | best plan is to chown youruser:www-data | 18:36 |
nobitanobi | by default | 18:37 |
nobitanobi | there is an index.html | 18:37 |
nobitanobi | and it is like this: -rw-r--r-- 1 root root 178 2011-06-25 19:01 index.html | 18:37 |
nobitanobi | so it's root and not www-data | 18:37 |
qman__ | yes, but notice it has world-read | 18:37 |
qman__ | so www-data can read it | 18:37 |
qman__ | as can everyone else | 18:37 |
qman__ | in order for apache to serve a page, it must be able to read it, and apache runs as www-data | 18:38 |
nobitanobi | ah I see | 18:38 |
nobitanobi | I usually get confused with users and groups | 18:39 |
nobitanobi | but I see I should be the owner, and the group should be www-data | 18:39 |
qman__ | doesn't have to be, but it's a good setup | 18:39 |
qman__ | that way if you want you can remove permissions to everyone on a file | 18:39 |
qman__ | most of the time not needed, but leaves that option open | 18:39 |
qman__ | and seeing files owned by www-data reminds you that they are visible though the web | 18:40 |
nobitanobi | :) | 18:40 |
nobitanobi | you are a good teacher | 18:40 |
nobitanobi | thanks | 18:40 |
=== oCean_ is now known as oCean | ||
=== JasonnAWAY is now known as Jasonn | ||
=== fenris_ is now known as Guest55070 | ||
=== Guest55070 is now known as ejat | ||
ejat | hi .. is there any package available for bind-sdb for ubuntu? | 20:36 |
jmarsden | ejat: I think not, there is a maintenance issue: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=118715 | 20:40 |
uvirtbot | Debian bug 118715 in bind9 "bind9: LDAP backend" [Wishlist,Open] | 20:40 |
ejat | owh .. that one with ldap .. | 20:40 |
ejat | i mean for sdb with postgresql ? is there any ? | 20:40 |
ejat | means .. anyone doing the patch with it .. | 20:41 |
uvirtbot | New bug: #802016 in dovecot (main) "package mail-stack-delivery (not installed) failed to install/upgrade: subprocess installed post-removal script returned error exit status 1" [Undecided,New] https://launchpad.net/bugs/802016 | 20:42 |
uvirtbot | New bug: #802034 in postfix (main) "package postfix 2.8.2-1ubuntu2.1 failed to install/upgrade: el subprocés s'ha instaŀlat el script post-installation retornà el codi d'eixida d'error 75" [Undecided,New] https://launchpad.net/bugs/802034 | 21:56 |
=== NG_ is now known as ng_ | ||
=== Jasonn is now known as JasonnAWAY | ||
=== ng_ is now known as NG_ | ||
keyboardtalk | How do I mount or connect to a network drive? | 23:37 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!