[00:04] SupaYoshi: did you read what you pasted? [00:04] I need to build php5 apache server on Ubuntu 16.04.2 LTS but PHP5 is deprecated from sources. Anyone have information on getting php5 on this box? [00:05] Cyb3r-Assassin: use 14.04 in a container/VM [00:05] Cyb3r-Assassin: or you can use ondrej's ppa, but it's not officially ubuntu (and only ondrej can support you with issues) [00:06] nacc: I dont have VM options. [00:06] or ability to build out a new image. [00:07] i would be fine with modifying sources.list with new ppa yes [00:08] perfect done thank you [00:53] can I mask the use of anything over 5.6? Having some problems with deps while building lamp [00:58] you may have better success using lxd to run 14.04 LTS tools instead.. [00:58] sarnold: +1 [01:30] hi there guys if i upgrade my linux ubuntu version with custom compile kernel do i lost the custom kernel.? [01:31] second question i tried to re compile the kernel of ubuntu server just increasing the time frequency but when i compile it its been two days already it still not done compiling , the status of compilation did not stop, is this really taking so much time and this long.? [01:35] guys anyone.? [01:36] ruben23: use 'make localmodconfig' to build a kernel just for your hardware [01:36] ruben23: the default kernels ubuntu builds have so many things enabled because you never know what hardware people have [01:37] a kernel build takes our buildds about five hours https://launchpad.net/~canonical-kernel-team/+archive/ubuntu/ppa/+build/12486778 so perhaps you've got slow hardware too? [01:42] sarnold: any chance we can speed it up.? [01:42] ruben23: make localmodconfig is a good start [01:42] just run that command.? [01:42] then be sure to run make menuconfig afterwards to set whatever variables you need differently than the ubuntu kernel's choices [01:43] the problem her eis we only need to change one thing timing frequency we just increase it thats all we need [01:44] but the compiling process is taking so much time [01:44] how many cores are you giving it? [01:45] sarnold: so i run make localmodconfig then make menuconfig after.? any guide you can send so i can push the process somehow [01:45] and threads did you tell make to use? [01:45] patdk-lap: no, just timign frequenzy to 1000HZ [01:45] more hz isn't going make kernel compile faster [01:46] it just makes task switching smoother [01:46] oh, you wanted to build with a higher hertz [01:46] isn't that dynamic these days? [01:48] ruben23: when you started the compile how many threads did you start up? [01:50] sarnold: i did not touch the threads, just the timing frequency [01:51] patdk-lap: yes what you mean dynamic.? [01:51] how to do that.? [01:51] what are you doing that needs 1000hz? [01:52] ruben23: when you run 'make -jN bzImage' what do you used for 'N'? [01:53] I like to use 0.1 :) [01:53] patdk-lap: what the heck does that do? :) [01:54] syntax error probably [01:55] the last time I built a kernel I used 'make defconfig' and 'make -j' with no arguments, and it finished in 48 seconds :) but I suspect that kernel would be too useless to do anything with.. [01:56] load average over 300.. [01:56] heh [01:56] I haven't built a kernel for awhile [01:56] I only built that one to test the machine, hehe [01:56] have built many illumos kernels, those take me about 45min on 6cores [01:57] well, that is more than a kernel, that is kernel + libc + basic tools [01:57] it gave me fond memories of 30-minute builds back in the pentium days.. [01:57] oh? I'm surprised they don't have a GENERIC that just works [01:57] generic? [01:57] yeah, like the BSDs [01:57] not sure what you mean [01:58] the BSD kernels usually hvae a bunch of configuration options same as linux, but they always recommend everyone to run the prebuilt GENERIC builds they provide [01:58] I remember recompiling freebsd kernels like mad [01:58] those never worked for me, freebsd only supported 2gigs of ram, and going over it, required extensive tuning [01:58] oww [01:59] cause the default table sizes where based on ram sizes, and they would just get stupid big, or be way too small [01:59] for illumos, it's cause of kernel patches I have made, or drivers I'm adding to it [01:59] 1024 processes should be enough for anyone! [01:59] hehe [02:00] oh cool, what have you done to illumos? :) [02:00] bcrypt work [02:00] testing a bunch of patches for other people [02:00] adding in pvscsi and vmxnet, though those are mostly offical now [02:00] endless hacking with lsi sas driver [02:01] i'm surprised that's not perfect by now; everyone uses that, right? [02:01] Used to do linux kernel work, but haven't touched it since 2.2 [02:02] did a crapload of hacking of my 2.0 and 2.2 kernels, complete memory and scheduler swap outs [02:03] back when that kind of thing made more of a difference, hehe [02:04] I would still enjoy a scheduler option [02:04] used strict priority levels [02:05] nothing on a lower level would run, unless all higher level niced programs where idle [02:05] having a runaway program would DOS your system, but [02:06] recommend, don't use firefox on it :) [02:06] but then, those where servers, and I didn't really have that issue [02:06] you -could- configure linux to do the same; put everything into sched_rr and remove the slack space that allows non-rt tasks to run even if rt tasks peg the cpu [04:04] Im using this ppa https://launchpad.net/~ondrej/+archive/ubuntu/php to give me php5.6 but I am having dep conflicts trying build a LAMP service that supports everything correctly. Any PPA I can add to provide everything else I need? [04:07] Or any option to have it build with apt-get install lamp-server^ as I need. [06:32] Good morning [06:52] Hmm "Setting up libc6:amd64 (2.19-0ubuntu6.11) ..." crashed with a reboot. oO [06:53] A "dpkg --configure -a" crashed also, the maschine does ja full reboot. :( [06:53] 14.04 [06:54] Any ideas? :( [06:54] Bodenhaltung: Grab a live-usb, chroot into the install and update from there. [06:55] Hmm, not really possible, its a virtual maschine (on openvz) [07:00] lordievader[m]: Can i solve the "blocker" so i can --reinstall libc6? [07:01] Bodenhaltung: Oh, that makes it easier. Mount the virtual disk in the hypervisor and chroot from there. [07:15] Bodenhaltung: hey I've got an idea.. it's a bit of a wild guess [07:16] Bodenhaltung: within the openvz, run "mkdir /run/systemd/system" [07:16] or "mkdir -p /run/systemd/system" [07:17] Bodenhaltung: http://sources.debian.net/src/glibc/2.19-18%2Bdeb8u9/debian/debhelper.in/libc.postinst/#L221 [07:33] Hello. I am ubuntu server for ftp services but as soon as I type "ifconfig" I cannot see eth0 or eth1 being listed; instead I just see enp0s3 and lo interfaces. Why? What's the matter? [07:34] Newer udev ;) [07:34] Hello [07:34] pankaj_: https://www.freedesktop.org/wiki/Software/systemd/PredictableNetworkInterfaceNames/ [07:35] lordievader[m]: Sorry Sir but I googled it and got the same article but the documentation seems to be very complex. Is their is no any straightforward reason and what should I do for my ftp server to assign IP address? [07:36] It is not very complex, basically the way the name is derived has changed. [07:37] Rather that something arbitrarily a consistent way based on where the network interface is. [07:37] The physical location of the nic on the motherboard. [07:37] The ways of assigning ip addresses has not changed, but instead of having an ethX name you have an enpXsY name. [07:38] lordievader[m]: I think that "yes" it is not complex. OK, so is their any way so that I can assign the IP address so to run my ftp Server. [07:38] Dchp does not cut it for you? [07:39] lordievader[m]: I didn't understood. [07:39] pankaj_: https://help.ubuntu.com/lts/serverguide/network-configuration.html [07:40] @lordievader sorry to interupt you on that but do you know the reason they changed that name usage? [07:40] lordievader[m]: Thanks. I will read the documentation and then. [07:41] microwav_: machines with a few NICs would randomly re-order the stupid things on reinstalls [07:41] microwav_: Read the documentation. The motivation is to make device naming predictable. [07:41] ah that clarifies, thnx! [07:41] microwav_: or deployments of a thousand 'identical' machines would have unexpected names roughly distributed randomly [07:41] microwav_: or people with multiple usb nics not knowing which is which.. [07:43] Or the networking not working because you transferred a disk to another exactly the same machine because the mac address was different. [07:43] sarnold lordievader[m] Its solved, i have reinstalled kmod, now it is working fine, thanks to you for point me in that way. :) [07:43] Network was configured for eth0, not for eth1 :P [07:43] Bodenhaltung: Good to hear :) [07:44] Bodenhaltung: kmod? [07:44] thnx Lordie [07:44] and sarnold [07:46] sarnold: Jepp, kmod:amd64, pure desperation. :D [07:49] Bodenhaltung: and that -worked-??? [07:52] lordievader[m]: Thanks. I think firstly I must read a good book to get started with networking environment in Linux. I had previously studied networking but can you suggest any good books so to get started in linux networking environmrnt? [07:53] Errrmm, I have gotten my knowledge mostly from online documentation and trial & error. [07:55] lordievader[m]: Any advice from you will be a lot to me. [07:55] I started with ora.com's TCP/IP networking book ~twenty years ago. If it's been kept updated over the years it might be alright today. [07:56] pankaj_: Grab a vm and experiment. Read documentation, try out what is said, see what it does and try to understand why things work the way they do. [07:56] If you -really- want to know your stuff, W Richard Stevenson's many books on TCP/IP are fantastic. Sadly they have probably not been updated over the years. [07:58] sarnold: I just want to know the networking environment basics in Linux. I just downloaded a Book "Linux Device Driver" by Jonathan Corbett. Can you suggest anything about the book [07:59] lordievader[m]: I had already installed it and now running ubuntu server but I think I must also understand all things that I am doing and not just type some code so that things workout. Therefore, I said abut any good book or documentation. [08:00] pankaj_: that's a fantastic book but it is all about writing kernel device drivers. If that's your goal by all means keep at it :D [08:01] sarnold: Yes, if it makes understand the reader to have some good grasp of linux kernel and networking then I will surely read this book. [08:02] sarnold: Can you suggest me any any book whose focus is all on networking in Linux environment? [08:03] mdeslaur: strongswan remerge for CVEs done [08:04] at least for me and some tests it works as before but safer :-) [08:04] pankaj_: this is entirely linux and entirely networking -- it's also extremely advanced material http://lartc.org/ [08:05] sarnold: OK. Firstly I must test the book by reading myself. Thanks. Talk to you later [08:05] pankaj_: have fun :) [08:05] pankaj: Thanks [08:05] pankaj_: also slightly outdated but a great start and ofen more a use-case/coding point of view that I liked was in "UNIX Network Programming" by Richard Stevens [08:06] sarnold: is that the same you recommended on a different title? [08:06] pankaj: I will also try it and then decide between the two. Well, the name seems to be very interesting. Thanks for telling. [08:06] sarnold: or did you mean the TCP/IP Illustrated series? [08:07] pankaj_: theres also the "adcanced" version of it [08:07] anyway we all seem to sellte that mister Stevens is a good way to start :-) [08:07] cpaelzer: OK. I will check it out. [08:07] I was lucky to get most "basic books" when they closed an IBM Library a few years ago [08:08] and basic concepts rarely change [08:12] cpaelzer: I meant the TCP/IP Illustrated series [08:12] UNIX Network Programming is good too but the entire second half of the book is a long-dead interface, right? [08:12] yep [08:12] I just always liked the "parcitical" start [08:12] cpaelzer: oh man, grabbing stuff from an ibm library, great idea :D [08:13] I've had more than I could move on my last move to a new place [08:13] handing out books all around [08:13] hahaha [08:14] when I moved into my current house I was thrilled that it already had something like 5m * 4 shelves... but it's not enough. I still have book piles. [08:14] And it was a nice pic when 4 studends filled up my crappy old car with books until the suspension stopped us to go on [08:14] rofl [08:14] wunderbar :D [08:15] sarnold: wow 5m*4 is impressive - these days I read like a snail crawling over the pages, I won't need more anytime soon [08:17] cpaelzer: funny thing, I read so much more on my phone now, I'm still working on harry potter und der order der phoenix for five or six weeks... my german's not quite good enough to plow through it even when tired. hehe. [08:19] sarnold: phönix :-) with ö [08:19] cpaelzer: my keyboard lacks those :) [08:19] it's vice versa for me only reading in english since a decade [08:19] damn actually two decades - oO getting old [08:19] hehe === clvx is now known as Guest35623 [10:44] Is there a reason for nginx package to install both fastcgi.conf and fastcgi_params, with the former having SCRIPT_FILENAME defined in addition to other params also present in fastcgi_params? [10:45] I mean, since there are uwsgi_params, proxy_params and scgi_params, why not just fastcgi_params, with SCRIPT_FILENAME, as default? [10:45] It's inconsistent and confusing to new users. [11:03] cpaelzer: thanks! === fallentr1e is now known as fallentree === soahccc_ is now known as soahcc === soahcc is now known as soahccc [14:05] Need to perform CPU benchmarks on Linux systems, recommended tools? [15:05] teward: --^ fallendtree's comments? [15:05] *fallentree [16:32] The server team CI bot owner might be interested in this bug I just reported about Launchpad. https://bugs.launchpad.net/launchpad/+bug/1695031 [16:32] Launchpad bug 1695031 in Launchpad itself "apport retracing service regularly encountering timeouts" [Undecided,New] [16:33] bdmurray: thx [16:35] powersj: Are you seeing these timeouts too? [16:35] <-- server team CI bot owner and no I have not seen evidence of them, so I will need to go look where something is complaining. [16:37] powersj: Okay, I'd be curious to hear about what you discover. [16:37] ok I'll let you know if I find anything [16:39] bdmurray: that didn't take long https://jenkins.ubuntu.com/server/job/simplestreams-ci-trigger/50799/console [16:40] that is a job to check for merge proposals that need reviews [16:43] powersj: Any idea when the bot first started encountering it? [16:43] bdmurray: I don't. I only have a days worth of run logs. I haven't been keeping an eye on streams project because it is usually low volume/nothing to do there :\ === Epx998- is now known as Epx998 [21:06] rbasak: I think LP: #1275495 should drop bitesize and server-next, agreed? [21:06] Launchpad bug 1275495 in monitoring-plugins (Ubuntu) "check_radius missing" [Medium,Triaged] https://launchpad.net/bugs/1275495 [21:08] nacc: agreed, thanks [21:09] rbasak: thanks [22:20] nacc: one last q, how can I inspect the change that went into 0.43ubuntu1.1 in xenial? [22:20] for pkgsel [22:22] dpb1: head to the launchpad page for the source package https://launchpad.net/ubuntu/+source/pkgsel [22:22] dpb1: couple of options [22:23] click on the version you want, then find the 'diff from 0.43ubuntu1 to 0.43ubuntu1.1 (799 bytes)' near the end [22:23] sarnold: ah yes [22:23] dpb1: what sarnold said, which should get you to https://launchpad.net/ubuntu/+source/pkgsel/0.43ubuntu1.1 and then http://launchpadlibrarian.net/265398320/pkgsel_0.43ubuntu1_0.43ubuntu1.1.diff.gz [22:23] * dpb1 waits for hopefully command line option [22:24] dpb1: or, locally, `pull-lp-source -d pkgsel 0.43ubuntu1; pull-lp-source -d pgksel 0.43ubuntu1.1; debdiff pkgsel_0.43ubuntu1.dsc pkgsel_0.43ubuntu1.1.dsc` [22:24] If only pkgsel were imported into git :-) [22:24] dpb1: or you can ask me to import it and then look at the git diff [22:24] rbasak: that was #3 :) [22:24] secret option #3 [22:24] OK, that is enough pestering. that's exactly what I wanted, thanks [22:25] * dpb1 dupes that bug now [22:25] dpb1: now, there is a separate question -- why wasn't this caught in testing? [22:25] dpb1: does that mean we don't have a written testcase for server iso for the possible choices here? [22:26] dpb1: i mean, it was caught pretty early, but still post-release [22:26] nacc: I'd like to explore that, yes [22:26] dpb1: oh it also may have been a bug in the merge, I see that now (based upon the prior xenial changelog entry) [22:29] hm [22:29] for that bit of archeology, I think I need the git import [22:30] s/need/want/ :) [22:30] dpb1: it's running [22:30] it was introduced in the runup to xenial for sure [22:58] dpb1: sorry, src:pkgsel showed me a bug in the importer (not super important, but breaks our orig tarball imports). I was looking at fixing it, but i'll generate the tree anyways so you can do archaeology :) [22:58] s/tree/repo/ === helohelp is now known as helpImStuck