| Freemon | hello | 05:33 |
|---|---|---|
| Freemon | did anyone try to install phantheon desktop on server 17 ? | 05:34 |
| === ashleyd is now known as ashd | ||
| lordievader[m] | Good morning | 09:35 |
| obinoob | Hi, need to set SET explicit_defaults_for_timestamp=1 but I can find a way, I'm on ubuntu server mysql 5.5.55 | 09:56 |
| sarnold | obinoob: this looks like it was introduced in 5.6.6 https://dev.mysql.com/doc/refman/5.6/en/server-system-variables.html#sysvar_explicit_defaults_for_timestamp | 10:00 |
| obinoob | sarnold: can I upgrade mysql without loose my data? | 10:01 |
| sarnold | obinoob: it looks like mysql-5.7 is in main in 16.04 LTS | 10:04 |
| sarnold | obinoob: I see many package upgrade failures every release -- it feels like the package management scripts are a bit brittle, so if you've done anything out of the ordinary to your configuration it might take some work to untangle it | 10:05 |
| sarnold | obinoob: and of course backups are always a good idea | 10:05 |
| obinoob | no i didn't did anything special to my system is pretty much stock | 10:06 |
| obinoob | I will follow this tutorial is based on digital ocean https://askubuntu.com/questions/750498/mysql-5-5-update-to-mysql-5-7 | 10:06 |
| sarnold | obinoob: if you do that you're on your own for security updates -- be sure to keep up with those every two months. | 10:07 |
| obinoob | sarnold: so whats my best shot here? | 10:08 |
| sarnold | obinoob: I'd be inclined to upgrade to 16.04 LTS. The worst of the bugs have probably been shaken out of it by now :) | 10:08 |
| obinoob | sarnold: well I've a lot of server configuration done DNS DHCP, OPENVP, APACHE, PHP etc not sure ... | 10:09 |
| sarnold | obinoob: if you'd rather use mysql straight from oracle, you can give those a try, but it may not be easy to return to using ubuntu-provided packaging | 10:09 |
| sarnold | aha, yeah, 16.04 LTS's php is php7. A huge amount of php software is still stuck in 5.3 days.. | 10:09 |
| obinoob | I'm running php7.1 here from ondreij I guess... | 10:10 |
| obinoob | anyway I will probably upgrade to a stable release in future but for now got to solve this ET explicit_defaults_for_timestamp=1 | 10:11 |
| sarnold | bed time here, good luck :) sorry I can't be more specific, I'm not sure about most of these things | 10:11 |
| sarnold | i'm happy enough to not need to tend mysql or php :) | 10:11 |
| obinoob | where are you? | 10:12 |
| obinoob | oceania? | 10:12 |
| obinoob | japan? | 10:12 |
| obinoob | lol | 10:12 |
| obinoob | ok thank you very much | 10:12 |
| sarnold | portland, oregon, usa | 10:12 |
| sarnold | nn :) | 10:13 |
| obinoob | ;) | 10:13 |
| === JanC_ is now known as JanC | ||
| supercool | Hello all! | 21:10 |
| supercool | I am running a Ubuntu server very fast and the other is just not. I was checking with top and I see one is using swap memory and the second is not. | 21:10 |
| supercool | My question is how do I turn swap memory on? | 21:11 |
| supercool | Is it even possible? | 21:11 |
| genii | !swap | 21:16 |
| ubottu | swap is used to move unused programs and data out of main memory to make your system faster. It can also be used as extra memory if you don't have enough. See https://help.ubuntu.com/community/SwapFaq for more info | 21:16 |
| jelly | supercool, does the second use none, or have none configured at all? | 21:24 |
| supercool | jelly: I already set it to 1GB | 21:45 |
| supercool | The thing is I have a processo running that in a few minutes fallows from 100% CPU to 17%... It used to be 17% all the time. | 21:45 |
| supercool | I think it is related to swap disk now. First there where no swap disk. | 21:46 |
| supercool | Now the disk gets filled to quick perhaps? | 21:46 |
| supercool | I set a 1GB swap and it rolds 20secs with full working CPU 100%, then it drops to 17% more or less. | 22:05 |
| supercool | Now I set a swap of 4GB and it took 1 minute with 100% CPU | 22:06 |
| supercool | Now it finally drop to 28% more or less. | 22:06 |
| supercool | I wonder if I can set something to recicle or something like that the swap partition | 22:06 |
| supercool | Maybe this application is just filling it | 22:07 |
| supercool | Any ideas please? | 22:07 |
| tomreyn | supercool: this sound slike you need to debug the application. | 22:41 |
| tomreyn | examining why swap grows rapidly on OS level is not really possible, since it's processes demanding it and that's all you'll find out. | 22:42 |
| supercool | tomreyn: is there a way to confirm this application is using too much swap? | 22:43 |
| supercool | just a basic level, as user | 22:43 |
| supercool | well, besides what I am doing of course | 22:43 |
| supercool | What else could be dropping down the CPU speed like that? | 22:44 |
| tomreyn | supercool: there is no reliable way to determine swap allocation by process, but the *2nd* answer here has something which *may* give a good enough hint https://stackoverflow.com/questions/479953/how-to-find-out-which-processes-are-swapping-in-linux | 22:48 |
| tomreyn | you may also wish to look at iotop | 22:49 |
| supercool | checking... | 22:50 |
| tomreyn | i.e. you need to find out what is your bottleneck. often the thing which is the boittleneck causes other perfomance indicators to change as well. but you need to identify the root cause. | 22:51 |
| supercool | The thing is this same application runs at two different systems | 22:52 |
| supercool | One runs it per hours all right | 22:52 |
| supercool | The other got stuck in 1 minute | 22:52 |
| tomreyn | high disk io can result in elevated cpu overhead (e.g. if you have software storage encryption). ram allocation spikes can cause high disk i/o due to swap configuration and and swapiness. | 22:52 |
| tomreyn | and the application work load is the same on both for sure? | 22:53 |
| supercool | Thats why I thought it could be something system related, you know.. | 22:53 |
| tomreyn | identical hardware? | 22:53 |
| supercool | tomreyn: yes | 22:53 |
| supercool | same loads | 22:53 |
| supercool | tomreyn: no, very similar in memory and disk I think. I could check that. | 22:53 |
| supercool | But somehow the second one got stuck! | 22:54 |
| tomreyn | you should. if hardware is the same, look at firmware. then look at hardware health, hwere possible (e.g. S.M.A.R.T. and the like) | 22:55 |
| supercool | how can I list the users list from the users listed on top command? | 22:55 |
| tomreyn | also check kernel messages / logs, i.e. /var/log/dmesg + /var/log/syslog | 22:56 |
| supercool | the top command is saying there are 11 users at one system | 22:56 |
| tomreyn | w, who, last | 22:56 |
| supercool | Ok | 22:56 |
| tomreyn | /var/log/auth.log also | 22:56 |
| * tomreyn bbl | 22:57 | |
| supercool | how about the memory being used | 22:58 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!