[01:36] I'm trying to install ubuntu server edition on a dell poweredge 2800 but I can't seem to get the disks listed when trying to install. There are about 7 hard drives present so I don't know how this is happening [01:36] Anyone ever handled one of these things? This is my first time [01:37] Originally when I tried to install from USB the only drive it listed was the cdrom, burnt the iso to it and now nothing is listed [01:40] do you need to reconfigure the perc card in some fashion? e.g. set it to hba mode or set up a raid array via bios or something similar? [01:44] I don't know what a perc card is. PXE is an option. I also don't know what hba mode is. I tried finding something to do with the RAID configuration in the BIOS but no luck [01:46] amazoniantoad: press ctrl-r (as in RAID) during boot. PERC is a storage controller (Dells' proprietary HW RAID) [01:46] tomreyn: doesn't display that as an option but I'll try that right now [01:47] bios revision a06 [01:47] utility mode, setup, pxe mode [01:47] there's a chance it might be prompting you while video doesn't work.. [01:47] servers are cranky like that [01:48] also, utility mode might get you there [01:48] It's still loading, very very little memory on this machine [01:48] some screen shots here https://www.dell.com/support/article/us/en/04/sln292433/dell-poweredge-no-boot-device-available-is-displayed-during-startup?lang=en [01:49] options so fuck [01:49] scsiselect utility [01:49] bmc setyo [01:49] setup* [01:49] so far* [01:49] wth? [01:49] Why did it autocomplete with a curse... [01:50] So it offers all of those options [01:50] then it boots into my installation media [01:53] enter setup, go to "integrated devices", see what "Embedded RAID controller" is set to [01:54] ...and which values are available for this option. [01:54] okay [01:56] os install mode is off [01:57] dont see embedded raid controller [01:57] okay maybe you just have none then [01:57] No what? [01:57] I know there are hard drives [01:57] I'm looking at them [01:58] are they plugged into a backplane? is the backplane connected to anything on the motherboard? [01:58] I'm going to have to take this thing apart to get a better look [01:58] Because there are big sheets of metal covering up spots [01:59] I'd *hope* the connectors for the backplane would be accessible with just removing the top.. [01:59] afterall someone might need to service the machine while it's still on the rails.. [01:59] I see embedded server management [02:00] This guy just left this server at this office and it didn't actually do anything. It ran some special version of windows but all he used it for was for employees to clock in [02:00] We have shifted away from that and now I have to repurpose this thing [02:01] I'll be back [02:01] I'm going to start taking this thing apart [02:02] "Embedded dual channel Ultra320 SCSI" .. heh, this is older than I expected ;) [02:02] :) [02:03] there is an add-on raid-controller, PDF page 6 of https://downloads.dell.com/manuals/all-products/esuprt_ser_stor_net/esuprt_poweredge/poweredge-2800_user%27s%20guide8_en-us.pdf explains how to locate it. [02:03] but since it's an add-on it makes me think it's not installe dby default. [02:03] server manuals https://www.dell.com/support/home/us/en/04/product-support/product/poweredge-2800/manuals [02:04] "raid key" [02:04] latest BIOS is A07, 07 Jul 2012 [02:04] cute, isnt it? [02:04] remind me to hug my supermicro when I head back downstairs :) sheeeesh [02:04] :) [02:05] latest BMC Firmware, v.1.83, A10, 07 Jul 2012 [02:05] latest Backplane Firmware, v.1.00, A01, 07 Jul 2012 (so probably no upgrade needed there, phew) [02:06] hah [02:17] 2 gb ram installed by default [02:18] but dell selles refurbished 256 MB DDR2 ecc upgrade modules for it [02:18] so you're safe. [02:18] oh man.. how much do they want for those? :) [02:18] dunno, no prices there. [02:18] uhoh, talk to sales staff territory, that can't be good. they have boat payments to make.. :) [02:19] :) [02:20] 15k rpm scsi drives [02:20] 36 GB :) [02:22] I wonder how throughput / latency on those compare to my spinning metal disks.. probably pretty favourably [02:23] 2x 1900W PSU [02:23] might draw a little bit more power than yours [02:25] insert impressed whistle sound here [02:25] that's incredible [02:31] amazoniantoad: so it's actually Ctrl-M to enter RAID controller setup according to https://www.youtube.com/watch?v=RtbstKvdhBs [02:31] for "mraid"? :) [02:33] or "make it work", not sure [02:33] Ctrl-E for BMC, also doesnt make a lot of sense, unless it stands for enlightenment. [03:03] any ZFS expertise around? I am running into this and wondering what is up.. https://askubuntu.com/questions/1054051/zpool-import-not-importing [03:10] dpb1: zpool import -d /dev/disks/by-id/ or something similar --- just plain /dev/ if you don't care about stable device names in the vdevs [03:10] sarnold: ya, same behavior with either -d /dev/disks/by-id or -d /dev [03:11] :/ [03:11] dpb1: dmesg? [03:11] * dpb1 looks [03:14] nothing in dmesg, :/ looking if there is a way to set debug, etc [03:14] dpb1: hmm. okay, maybe make a new directory, and copy the exact device node into that directory.. [03:17] you mean, like mknod? [03:18] that ought to do it too.. but I thought cp would do that for you? [03:19] oh man. cp tries to *read* the thing. ew. [03:19] yeah then mknod. heh. [03:20] ok, did cp -dp [03:20] same [03:20] no pools available to import [03:23] hrmph [03:25] dpb1: time to head to #zfsonlinux and hope dehacked is around.. [03:27] asked [03:27] :) === tikun is now known as sikun [05:03] anyone can help me? i am having problem installing full gnome desktop on ubuntu 18.04 server [05:03] the command sudo apt install tasksel taking forever to finish [05:03] cleaned installed on a VM [05:03] installation and network config is done. i can log-in and now wanted to install gnome desktop [05:23] update: i am following this thread https://askubuntu.com/questions/53822/how-do-you-run-ubuntu-server-with-a-gui [05:23] it sas to do this "sudo apt-get update" and "sudo apt-get install ubuntu-desktop" [05:24] the first one sudo apt-get update is taking long time. is there anyway to do it faster?/ [05:38] qwebirc52334: why do you install a gui on a server? [05:39] to do works conveniently without remembering commands [05:40] if I had a GUI, browsing though the settings and files would have been easier for me [05:40] it effectively prevent that you learn your way around the terminal, but you really should if you want to manage servers regularly, since once you do, you will work much better, faster, more effectively. running a GUI also consumes system resources you could better use for other things. and last but not least, it increases the attack surface. [05:41] you could use text based GUIs such as ncdu which don't depend on X [05:42] your choice - i just would not recommend it. if you still would like to proceed, show the output of the apt-get update command. [05:42] !pastebinit | qwebirc52334 [05:42] qwebirc52334: pastebinit is the command-line equivalent of !pastebin - Command output, or other text can be redirected to pastebinit, which then reports an URL containing the output - To use pastebinit, install the ยซ pastebinit ยป package from a package manager - Simple usage: command | pastebinit [05:43] i understand. for example, right now I am having issue with very slow apt-get and I dont know why. my internet is very fast, and do not have any issue with windows servers at all. i have seen that I might have to change repository, how do I do that? if i had a GUI, i could probably find it out by myself [05:44] without a GUI, you can find out yourself as well, you just need to search the web a little. these tasks have been done and documented many many times before, so you're not alone. [05:45] sure, it will take a bit longer than doing it on a gui the first and second time, but by the third time you will start saving time. [05:45] and get much better error output. [05:45] !terminal | qwebirc52334 this is for later, to get you started [05:45] qwebirc52334 this is for later, to get you started: The linux terminal or command-line interface is very powerful. In Unity or GNOME, search the dash for "terminal" and press ENTER. Other desktops: Applications -> System Tools -> Terminal (MATE), K-menu -> System -> Konsole (KDE), or Menu -> Accessories -> LXTerminal (LXDE). Guide: https://help.ubuntu.com/community/UsingTheTerminal [05:46] ok, I am following.. [05:46] see the link 'Using the terminal' there [05:46] for now, i can guide you at least to find out why apt is slow [05:46] which ubuntu version is this, do you know? if not, run: lsb_release -ds [05:47] oh you said 18.04, sorry. [05:47] i dont have any error now.......but its very slow i can say. right now its stuck at 18% [11 sources 0 B/1,098 B 0%] [05:47] its 18.04 amd64 [05:48] okay, can you run this in a second terminal window: apt-cache policy | pastebinit - [05:48] this should return a http address, or an error message saying you dont have pastbeinit installed [05:48] ok, sorry for asking but how do I open second terminal? [05:49] are you connected to the server using ssh? [05:49] i am connected to console [05:49] so you have keyboard and mointor attached to it and are sitting necxt to it? [05:49] its a VM and I am directly connected to display [05:49] yes [05:50] okay, which virtualization is this? [05:50] VMware [05:50] a pity, i dont know enough about it. [05:50] esxi 5.1 [05:50] i can use my keyboard and mouse to inout to the VM console [05:50] i see, you can install an ssh server next time, then access the system via ssh. [05:51] this allows you to connect to it multiple times and have multiple sessions open at once. [05:51] there are better ways for this, but this is too advanced for now. [05:51] ok. may be after this is done or probably errored-out.. [05:52] i think i can connect to to the server by ssh [05:52] at how many % is it now? [05:52] 18% [05:52] let me open putty to connect to it [05:52] okay, just cancel it then. press ctrl-c [05:52] ok [05:52] oh you can connect with ssh already? [05:52] cancelled [05:52] putty is an ssh client [05:53] i didnt try before but i think i can connect to ssh. anyway, i cancelled the apt-get command [05:53] and thanks thanks for your helping me..... [05:53] okay, let's test internet access: type this, press enter, and tell me what it says for % packet loss on the bottom: ping -c 3 abc.de [05:54] ok [05:54] you're welcome [05:55] 3 packets transmitted, 3 received, 0% packet loss, time 7181ms [05:56] rtt min/abg/max/mdev = 283/285/287/1.670 ms [05:56] so thats the statistics [05:56] ok, that's pretty long, wher ein the world are you? [05:56] i am in Guam [05:57] ok, maybe then it is fine, i do not know how well the internet infrastructure is there [05:57] the internet speed that I usually get from our Windows servers is around 10+MB (mega byte) download [05:57] and germany where this server is is far from you. [05:58] its a 150Mbps connection [05:59] ok. is there anyway to choose a nearest package server? [05:59] okay, that's very good if it's per second, i was just referring to latency, not bandwidth. [05:59] okay, so let's see what is up with apt [05:59] right, we should do this [06:00] can you try to connect to the server via putty? [06:00] ok [06:00] if this works out then yxou can just copy and paste the commands i type, andwhat they return, this would be easier for you [06:00] i assume you had to type the statistics you provided above? [06:01] yes, logged in through putty [06:02] yes, u r right [06:02] very well, so what does this return? apt-cache policy 2>&1 | pastebinit - [06:02] now it will be easy [06:03] http://paste.ubuntu.com/p/Mss2khCF37/ [06:04] also, if you will come here regulalrly, i recommend you install an IRC client instead of using the website to login to the chat. personally i like the "hexchat" IRC client which is available on the major desktop operating systems [06:05] hmm, i see. well, I will try to get an IRC client. what client do u recommend? [06:05] see what i just wrote ;) [06:05] oh, got it [06:06] next, please post: ping -c5 archive.ubuntu.com 2>&1 | pastebinit - [06:07] ssh not responding......pls give me a moment [06:09] http://paste.ubuntu.com/p/TQjcJdBXQt/ [06:10] hmm, maybe soemthing is not right with the VM, let's check this: dmesg | pastebinit - [06:10] the trailing '-' is not actually needed [06:11] http://paste.ubuntu.com/p/kTkHqs3m9x/ [06:12] you can remove the virtual floppy if esx can do this. you will never need it. [06:13] yes [06:13] can you show: pastebinit <( free -m; df -h; ) [06:13] ok, i joined hexchat as sunshine5 [06:14] * sunshine5 :No such nick/channel [06:14] maybe you connected to a different IRC network. we are on Freenode. [06:14] sorry, rf53 [06:15] oh ok :) [06:15] hi tomreyn, this is me [06:15] hello hello. [06:15] see if you like this better [06:15] rf53: this is what a hilight looks like on hexchat [06:16] yes, better than webclient [06:16] can you show: pastebinit <( free -m; df -h; ) [06:16] http://paste.ubuntu.com/p/QmbjKgm7QS/ [06:18] what does "uptime" return? its just one line, post it here directly [06:18] i need to shutdown server to remove floppy drive [06:18] 06:18:32 up 1:06, 4 users, load average: 0.00, 0.00, 0.00 [06:18] okay, i'll wait for it to reboot [06:18] ok [06:19] Good morning [06:19] hi lordievader [06:19] i wonder why its showing 4 users [06:19] this is only me connected [06:19] w [06:19] ^ may tell [06:19] Hey tomreyn [06:20] you probably have 4 sessions open [06:20] How are you doing? [06:20] hmm. could be [06:20] ok, removed floppy drive and server is coming up [06:21] lordievader: fine, thanks :) rf53 + i are just looking into speeding up his apt. he's got http://archive.ubuntu.com configured right now but it's pretty slow from Guam. i'm reading that we could use "deb mirror://mirrors.ubuntu.com/mirrors.txt" instead [06:22] rf53: Do you have a slow internet connection? [06:22] rf53: give me a second, i am composing a new confguration file for you. [06:22] Slow DNS, etc? [06:23] hmm right we didnt check dns, it could be that [06:23] i dont have slow internet connection [06:24] i am using 150 Mmps corporate connection [06:24] from our windows servers, internet is very fast, like 10+MB download from microsoft repository [06:24] dont take it as an insult, we're just trying to understand what is making apt slow for you currently. [06:24] i understand [06:25] :) can you run: "systemd-resolve --statistics 2>&1 | pastebinit" [06:25] we have 100+ other servers, most of them ar windows and internet speed is not an issue [06:26] it could be soemthing with this VM, but since I am not a expert, not sure how to troubleshoot [06:26] time systemd-resolve archive.ubuntu.com 2>&1 | pastebinit [06:26] ok, running command... [06:26] rf53: What is the output of 'for i in {0..10}; do /usr/bin/time -v dig +short archive.ubuntu.com 2>&1 |grep wall; done'? [06:26] http://paste.ubuntu.com/p/z7fkkwrSFm/ [06:27] for i in {0..10}; do /usr/bin/time -v dig +short archive.ubuntu.com 2>&1 |grep wall; done [06:27] Elapsed (wall clock) time (h:mm:ss or m:ss): 0:05.96 [06:27] Elapsed (wall clock) time (h:mm:ss or m:ss): 0:00.00 [06:27] Elapsed (wall clock) time (h:mm:ss or m:ss): 0:00.00 [06:27] Elapsed (wall clock) time (h:mm:ss or m:ss): 0:00.00 [06:27] Elapsed (wall clock) time (h:mm:ss or m:ss): 0:00.00 [06:28] Elapsed (wall clock) time (h:mm:ss or m:ss): 0:00.00 [06:28] Elapsed (wall clock) time (h:mm:ss or m:ss): 0:00.00 [06:28] Elapsed (wall clock) time (h:mm:ss or m:ss): 0:00.00 [06:28] Elapsed (wall clock) time (h:mm:ss or m:ss): 0:00.00 [06:28] Elapsed (wall clock) time (h:mm:ss or m:ss): 0:00.00 [06:28] Elapsed (wall clock) time (h:mm:ss or m:ss): 0:00.00 [06:28] whoops, too much output here [06:28] sorry [06:28] np ;) [06:28] time systemd-resolve archive.ubuntu.com 2>&1 | pastebinit [06:29] http://paste.ubuntu.com/p/bhcyxSs6b5/ [06:29] real 0m2.662s [06:29] user 0m0.147s [06:29] sys 0m0.033s [06:30] That first request is quite slow. [06:30] ok [06:30] How is your dns set up? [06:31] it's ubuntu server in an esxi 5 VM [06:31] may be something with network configuration...i used ip address from our server address pool..like other servers [06:31] *18.04 [06:31] For comparison my first request (with an empty resolver cache) takes 0:00.01. [06:31] yes [06:31] esxi 5.1 [06:31] we have local dns server [06:32] Could you pastebin the output of `for i in {0..10}; do /usr/bin/time -v dig @8.8.8.8 +short archive.ubuntu.com 2>&1 |grep wall; done` please? [06:32] same LAN [06:32] lordievader: btw. rf53 is here (Guam): https://en.wikipedia.org/wiki/Guam#/media/File:Guam_on_the_globe_(Southeast_Asia_centered)_(small_islands_magnified).svg [06:33] but there seem to be undersea cables going there. [06:33] That is probably not going to help for latencies ๐Ÿ˜‰ [06:33] yes, its a small US territory [06:33] http://paste.ubuntu.com/p/Z7KGPrmkM5/ [06:34] That looks better. [06:34] I'd switch to Google DNS. Or, even better, setup a caching resolver wich forwards to Google. [06:35] so you're convinced it's dns? [06:35] i agree it seems to be base don your tests [06:35] Not convinced, but it plays its part. [06:36] maybe the local resolver is just a bit busy [06:36] how to configure the resolver properly on 18.04 ? [06:36] i really dont know yet [06:36] ok. we are in a LAN and not sure how to change/forward to Google DNS from Ubuntu [06:37] rf53: can you: pastebinit /etc/resolv.conf [06:38] I typically do not have resolvconf installed and can just make those changes to `/etc/resolv.conf`. [06:38] i think resolvconf is installe dby default though [06:39] http://paste.ubuntu.com/p/bVRzCnxpSW/ [06:39] Quoting 18.04 release notes: Default DNS Resolver: The default DNS resolver is now systemd-resolved. The standard /etc/resolve.conf file is now managed by systemd-resolved, and configuration of the resolver should be done in /etc/systemd/resolved.con [06:40] f [06:40] rf53: pastebinit /etc/systemd/resolved.conf [06:40] http://paste.ubuntu.com/p/BFzZWZH5Pb/ [06:41] hmm, okay then let's edit this file. do you know nano? it's a command line file editor. [06:42] do i need to install it on server? [06:42] it may already be installed. can you run:: sudo nano /etc/systemd/resolved.conf [06:43] yes, its here [06:43] we want to make line 15 of http://paste.ubuntu.com/p/BFzZWZH5Pb/ say this: DNS=8.8.8.8 [06:44] when done editing, press ctrl-x [06:44] and press enter to save it [06:45] will it be #DNS= or DNS= [06:45] do i remove # [06:45] no # [06:45] remove it [06:46] done, saved [06:46] now: sudo systemctl restart systemd-resolved.service [06:46] then run "apt-get update" again and see if its faster [06:46] And then test the latency again. [06:46] ok, doing... [06:47] also run this again: for i in {0..10}; do /usr/bin/time -v dig +short archive.ubuntu.com 2>&1 |grep wall; done [06:48] running apt-get update [06:48] let me open another ssh [06:48] :) very well [06:49] http://paste.ubuntu.com/p/PdvnqX5CMJ/ [06:49] please also show (just one line): date- u [06:49] can you say whether apt-get is faster now? [06:50] Wed Jul 11 06:50:12 UTC 2018 [06:50] thanks [06:50] apt-get seems again stuck at 18% [06:51] 18% [11 Sources 0 B/1,102 B 0%] [06:51] it must have trouble reaching one of your apt mirrors: http://paste.ubuntu.com/p/Mss2khCF37/ [06:52] bot security.ubuntu.com and archive.ubuntu.com are pretty standard, though. [06:53] ok, so I think I just need to wait for it to finish [06:54] is there anything like changing mirros to force it download from closest mirror? [06:54] yes [06:55] if you cancel apt (ctrl-c) and run this instead, it should tell us why it is taking so long: sudo apt-get -o 'Debug::Acquire::http=1' update [06:55] ok, do u think it might help? [06:55] ok [06:55] but we can also edit your /etc/apt/sources.list now, as you prefer [06:55] ... to maybe get you a faster mirrror. [06:57] with the command, looks like it was fetching the files. but again stuck at 18% [06:58] i am not sure what nearest mirror my location has [06:58] can you paste the last 15 lines of output to http://paste.ubuntu.com ? [06:59] any US or Australia Mirror? [06:59] probably australia, i think [07:00] https://paste.ubuntu.com/p/g56jhDFtW6/ [07:01] https://paste.ubuntu.com/p/NjD4dzrGH2/ [07:03] hmm okay, i was wroing, it doesn't help ;) [07:03] let's just update your sources.list [07:03] ok [07:03] you have ubuntu backports configured, do you need them? [07:04] no [07:04] ok, i'll remove them since they are not well supported. [07:04] ok sure [07:06] give me another 1 minutes [07:06] 2 [07:07] you are helping me, and I am grateful. take your time [07:09] cancel apt if it's still running, then run: [07:09] ok [07:09] cancelled [07:09] sudo mv /etc/apt/sources.list /etc/apt/sources.list-disabled-20180711 [07:09] done [07:10] wget https://pastebin.com/raw/pgBBUTXB -O - | sudo tee -a /etc/apt/sources.list [07:10] this will download the sources.list configuration file for APT which I posted to this location, and set it up for you. [07:11] ok, done all [07:11] please run "sudo apt update" afterwards, and let me know whether it works better [07:11] ok [07:13] looks like much faster.....so far [07:13] and not stuck [07:13] from AU [07:14] good. it may still get stuck sometimes in the future when you rnu it. if this happens, either just cancel and run it again (it will use different mirrors then and should not get stuck again), or wait. [07:14] ok, got it [07:15] now, do you still want to install ubuntu desktop? ;) [07:15] so apt-get update does download and install both? [07:15] no, it just uipdates the information on available packages, it doe snot install them ,yet [07:16] i think i will give some time to learn commands [07:16] you can also use 'apt' now instead of 'apt-get' (except when you write shell scripts with it) [07:16] apt is a little mor euser friendly [07:17] if you run "apt update" now it will tell you how to geta list of pending updates. [07:17] to install them, run: "apt full-upgrade" [07:17] And `apt` gives progress bars ๐ŸŽ‰ [07:17] so its like listing all pckages that has updates available. so to install update do I need to use apt-upgrade command? [07:18] to install them, run: "apt full-upgrade" [07:18] i will use apt then. progress bar i like more [07:18] ok [07:19] on my screen, though it seems working, its still says 18% [07:19] Get:11 http://mirror.tcc.wa.edu.au/ubuntu bionic/main Sources [829 kB] [07:19] 18% [8 Sources 0 B/1,113 B 0%] [11 Sources 0 B/1,098 B 0%] [07:19] at very last of lines, its 75.8 kB for all lines [07:19] so that's still the old apt command? you didn't cancel it yet? [07:20] not yet [07:20] well you can do so now [07:20] 75.8 kB is not THAT much ;) [07:20] ok, so let me do this [07:22] i ran "sudo apt update", seems same output except the last line is yellow color [07:22] now, it connected to NZ server [07:23] anyway, i will just wait and once its done will run "apt full-upgrade" [07:24] rf53: it will vary, picking the servers returned at http://mirrors.ubuntu.com/mirrors.txt at random (this file loos different for you than it does for me, since the server returns the best apt mirrors for your region) [07:25] got it [07:25] so since it picks different servers at random on the next run, if "sudo apt update" gets stuck, i would just cancel and run it again. [07:25] but that's up to you ;) [07:26] the problem is to know if its stuck or not, because the lines are moving.... [07:26] okay, if the output keeps changing then i guess it's fine [07:26] but its the same lines coming again and again [07:27] let me give u some recent lines in pastebin [07:27] they may look very similar, or even identical when it actually does somethign different. [07:27] ok [07:27] also, if you would like me to explain any of the commands we ran above, please let me know. [07:28] https://paste.ubuntu.com/p/w4rTycnJfy/ [07:30] so if "apt update" only lists updates needed, and not even does download, why is it taking so long? shouldn't it be done in just a few seconds? [07:31] Normally, yes. [07:32] its still 17%. let me cancel and run it again [07:32] hmm yes this doesnt look good [07:32] how can i check disk space info? sorry for asking [07:32] df -h [07:33] we did this before, you have space left [07:33] There are no firewalls at play? [07:33] http://paste.ubuntu.com/p/QmbjKgm7QS/ [07:34] oh ok [07:34] we have firwalls, but its not an issue for our windows severs [07:35] lines 1-3 are about memory in megabytes (free -m), the rest is the output of 'df -h' (disk free in human readable output) [07:36] when i traceroute to you from a couple locations around the world it always goes through los angeles, usa [07:36] so maybe we should actually use US mirrors [07:36] yes, we can try that [07:37] usually we have better traffic with US since its an US Territory [07:38] you can create another backup copy of the apt sources.list file: sudo cp -p /etc/apt/sources.list /etc/apt/sources.list-disabled-20180711-2 [07:39] created another backup [07:41] and then: sudo rm /etc/apt/sources.list; wget https://pastebin.com/raw/VhLc8Uv3 -O - | sudo tee -a /etc/apt/sources.list [07:42] done [07:42] and then: echo | tee -a /etc/apt/sources.list [07:42] this is just to add a blank line to the end, i forgot this [07:42] and finally: sudo apt update [07:43] sudo echo | tee -a /etc/apt/sources.list [07:43] tee: /etc/apt/sources.list: Permission denied [07:43] sorry, this: echo | sudo tee -a /etc/apt/sources.list [07:44] sudo apt update [07:44] Err:1 http://us.ubuntu.com/ubuntu bionic InRelease [07:44] Could not resolve 'us.ubuntu.com' [07:44] Err:2 http://us.ubuntu.com/ubuntu bionic-security InRelease [07:44] Could not resolve 'us.ubuntu.com' [07:44] Err:3 http://us.ubuntu.com/ubuntu bionic-updates InRelease [07:44] Could not resolve 'us.ubuntu.com' [07:44] uuh, my bad [07:44] do i need to update DNS again? [07:45] no, i just made a mistake, the server name is wrong [07:45] ok [07:47] to fix this: sudo rm /etc/apt/sources.list; wget https://pastebin.com/raw/Cehw3vSV -O - | sudo tee -a /etc/apt/sources.list [07:47] done [07:47] please also post: sudo ls -l /etc/apt/sources.list.d/ 2>&1 | pastebinit [07:48] and finally: sudo apt update [07:48] yes off course. its running ... [07:48] taking time at 14% [Waiting for headers] [07:49] 14% [4 Sources 0 B/1,104 B 0%] [07:51] hmm, something must be wrong with either oyu network or the esxi or the ubuntu installation [07:51] chances are oyu have double NAT and that gets in the way [07:52] with packets not getting properly rewritten or whatever [07:52] hmm [07:53] Perhaps doing a packet capture can give some insight. [07:53] you'd need to do a packet trace to analyze this further [07:53] ok, i will do that [07:53] i will also try to change ip address [07:54] thanks for your help [07:54] i will log-out now and try to do some troubleshooting myself [07:56] http://paste.ubuntu.com/p/rcr3DzqWKX/ [07:57] this is output of "sudo ls -l /etc/apt/sources.list.d/ 2>&1 | pastebinit" [07:58] okay, thats fine [07:58] i mean the output, and good luck with the touble shooting [07:59] ok. thank you tomreyn [07:59] it could also be a misconfigured firewall blocking the requests or, more likely, packets reutnrned for your requests [07:59] i suspect the tcp 3-way handshake doesnt get fully transmitted [08:00] you're welcome. [08:00] if you wanted to reproduce this on y another system, set up another VM on the same esxi server with similar network configuration and i suspect it will produce the same issue. [08:01] ok, what I will do it create another VM on other ESXi node and see if that has any difference [08:01] it could also be firwall if the packets transmission is different than windows [08:02] tcp should be universal, but more package translations can cause more problems. [08:02] good approach, good luck. [08:03] and if there is any rule in firewall [08:03] ok [08:03] thanks and bye === LaserAllan__ is now known as LaserAllan === LaserAllan is now known as laserallan === laserallan is now known as LaserAllan_ === RoyK^ is now known as RoyK_Heime [10:49] b move 8 [10:49] hello, I installed a fresh 18 server system, configured the eth interface to use dhclient at install time. How can I see what DHCP server provided info to the system? [10:50] anddam: can you rephrase that question? not sure I understand what you're asking. [10:55] blackflow: sure,I configured the eth interface on a system to use dhclient, it gets configured but I'd like to see which DHCP server provided the info [10:55] I read /var/lib/dhcp/ should have leases files [10:55] it does not on this system [10:56] on another similar system (these are virtualbox VMs with a bridged network interface) I do see the .leases file [10:57] the system showing the .leases file is an Ubuntu 16 and I configured /etc/network/interfaces, on this new 18 I see configuration moved to netplan [10:59] in /etc/netplan I see https://gist.github.com/adab4d/51cfd2d458602e439e8f1b08a7720b06 [11:01] it could also be that my assumption is wrong, how do I check that my enp0s3 has been configured via DHCP? [11:01] blackflow: ^ [11:04] anddam: netplan is just configuration abstration, it's using networkd as backend on servers, and NM on desktops. according to your post, it's configuring enp0s3 for dhcp. also cloud-init is mentioned, and that's where my knowledge about it stops, I don't use cloud-init [11:05] anddam: iirc, leases and dhcp activity should be logged in the journal [11:05] anddam: dhclient will log DHCPREQUEST and DHCPACK [11:13] blackflow: what service? [11:14] blackflow: as I wrote I think I configured for dhclient but I may be wrong, I would like to confirm that [11:14] sstemd-networkd[825]: enp0s3: DHCPv4 address 192.168.1.22/18 via 192.168.0.1 [11:14] correct server [11:14] thanks [11:15] anddam: right so you configured for DHCP, not specifically dhclient. that's systemd-networkd doing the work there. [11:43] oh, like dhclient is a specific DHCP client, but systemd-networkd can do the job as well? [11:44] hello all [11:44] anddam: iirc yes, it doens't need dhclient [11:44] how to get php pthreads mod with php5 on 14.04 lts ? [11:44] hello [11:46] Haris: like, for threaded PHP? I think you can't without rebuilding whole PHP [11:46] why do you need it btw? [11:48] is it possible via pecl mod install ? [11:48] php app that's scaling to require threading [11:48] its' not a module, it's whole interpreter being ZTS [11:49] is there a less radical solution for 14.x [11:49] PHP is not thread-safe in itself tho', so even with that, you'll likely have issues [11:49] it's not ubuntu specific, but PHP specific [11:51] don't the main repo or ppa pkgs have it pre-built somewhere ? [11:52] official repos don't. there could be a PPA but those aren't supported here. [11:52] hmm [11:52] where does one go for ppa [11:52] Haris: that said, see if Ondrej has something: https://launchpad.net/~ondrej [11:52] hmm [11:53] Haris: I'd do all this in a dedicated container btw. If you break your system, you get to keep the pieces. [11:53] lol [11:53] makes sense === apw_ is now known as ap === ap is now known as Guest79801 === Guest79801 is now known as apw [11:57] blackflow: is this sound right? https://gist.githubusercontent.com/hadifarnoud/4e5e3d1d2f977de2a337e43382ed726a/raw/8be2eb9717c4ce7b02da006201181cb701b85278/cloud-init.cfg [11:57] I think I should follow their instruction in comment [11:58] hadifarnoud: you don't need "network" or "netmask" parts, and 255.255.255.255 is invalid anway. use CIDR notation for the IP, eg. address 138.201.116.62/24 (if it's /24) [11:59] in fact not even broadcast is needed. only address and gateway (which you're missing) [12:00] it's 32 [12:00] hadifarnoud: the inet6 also looks wrong, I wouldn't use ::0 in that subnet. [12:00] hadifarnoud: no it's not. /32 means single IP but that's not your subnet. [12:00] my v6 is 2a01:4f8:1c17:5d80::/64 [12:01] is it though? that's all zeroes in that subnet (due to :: at the end) [12:01] ipv6 isn't about individual addresses but about subnets. you're assigned a /64 subnet and have to define addresses in it yourself [12:03] like this one? 2a01:4f8:1c17:5d80:78:47:223:238 [12:03] IPv6 is so confusing to me [12:03] hadifarnoud: looks okay if that's what you want. [12:04] I thought I assign a subnet and all IPs in that subnet works [12:04] not automagically no [12:04] do you know how many address that would be? :) [12:05] millions :D [12:05] 2^64 more precisely [12:06] I found my gateway via `ip route | grep default` [12:06] that should be ok, right? [12:06] can the pecl pthreads implementation help on 14.x lts ? [12:06] Haris: yes if that's a currently working config. it's also listed in your Robot (hover on the ipv4) [12:07] taht one went over me [12:07] that+ [12:08] sorry that was for hadifarnoud :) [12:08] Haris: as for pecl pthreads, I don't think so, you need ZTS in the interpreter [12:08] I mean try it, but I doubt it. [12:10] does it look fine now blackflow ? https://gist.github.com/hadifarnoud/4e5e3d1d2f977de2a337e43382ed726a [12:11] hadifarnoud: no. is that a hetzner CX VPS? that gateway looks like it. if so, you don't define public IPs on the iface. you also have two conflicting stanzas for eth0, one dhcp and one static. [12:12] blackflow it is a VPS. I think it does not persist so if anything goes wrong it's easier to just reboot, right? [12:12] hadifarnoud: but which VPS... Hetzner has different network setups between VX, CX and Cloud lines [12:12] CX11 [12:13] hadifarnoud: well, the CX line can have only one IP, 172.31.1.100, always, it's routed based on MAC from their side. you don't define public IP on it. [12:14] there is a float IP [12:14] hadifarnoud: not for CX [12:14] I have a floating IP on this server [12:14] then you'll have to ask Hetzner how to set it up. [12:14] afaik CX don't have it. new cloud servers do. [12:15] interest based money model is killing economies world wide. pushing human being(s) to reduce and reduce [12:15] isn't CX new cloud servers? that's what I have blackflow [12:15] it is not. [12:15] I currently did make it work by doing `ip addr add 78.47.223.238/32 dev eth0` [12:16] CX have no public IPs, only 172.31.1.100, always and only that. [12:16] so if your server has a non RFC1918 address and it works, that's not a CX VPS [12:16] I'm confused. how is it working now? [12:16] by not being a CX server [12:18] it clearly says that in dashboard. CX11 [12:18] this is slightly offtopic here though, you'll have to ask Hetzner for setup specific to their network. meanwhile, as for those e/n/i stanzas: write the IP in CIDR format. the gateway is in the same subnet, can't be RFC1918 then. you had conflicting entries for eth0 with dhcp and static at the same time. [12:19] well if that's CX and you have a non-RFC1918 IP, then they changed something significantly, or you aren't using what you think you are. At any rate, ask Hetzner for support with that. [12:19] https://gist.github.com/hadifarnoud/4e5e3d1d2f977de2a337e43382ed726a see my comment. there is a screenshot there [12:20] how come gateway is not in that subnet? you're right. maybe `ipconfig | findstr /i "Gateway"` isn't right [12:20] is pthreads supported on xenial or bionic or same for them also ? [12:20] php pthreads [12:21] ondrej says no pthreads before 7.0.x [12:21] hadifarnoud: oh I see what they did..... they started calling the new cloud servers as CX, and they dropped the old CX servers [12:21] that means its experimental on xenial [12:21] but its already available on centos [12:21] via a third party repo [12:21] ip route | grep default [12:22] yeah [12:22] Haris: php zts is not officially supported on Ubuntu, so this is becoming slightly offtopic. like I said, if you want that, you're on your own, use a PPA or build from source, good luck ;) [12:22] so, the gateway is correct? [12:22] hadifarnoud: I don't know. afaik their cloud servers don't use RFC1918 for anything [12:23] _really_ open a ticket for them and ask for proper network config..... OR, reinstall from their original ubuntu images [12:23] if you want a floating IP, convert from dhcp to static, have two stanzas, one for each IPv4 [12:23] btw, their floating IP is really manual work, it doesn't auto failover. [12:25] how about this? https://gist.github.com/hadifarnoud/4e5e3d1d2f977de2a337e43382ed726a [12:27] hadifarnoud: :0 and :1 are not needed these days any more (wiht iproute anyway). that gateway doesn't look right, but if they say it is, then it is. /32 is not your subnet. [12:27] I'll ask for the gateway [12:28] so I can just use the same interface? https://gist.github.com/hadifarnoud/4e5e3d1d2f977de2a337e43382ed726a [12:28] hadifarnoud: check the console, all networking info should be listed there [12:30] unfortunately there is nothing more than server IPs and reverse DNS in Hetzner Console. [12:31] hadifarnoud: well launch another cloud instance and see how it's set up. you pay them by the hour anyway [12:31] (using their official ubuntu images, I mean) [12:32] ok [12:34] rbasak: hi, good morning/afternoon [12:34] rbasak: if you have a moment, could you please push the upload tag for https://code.launchpad.net/~ahasenack/ubuntu/+source/libapache2-mod-perl2/+git/libapache2-mod-perl2/+merge/348780, after you take a quick look? [12:34] cpaelzer +1'ed it [14:21] jamespage: btw i've been making python(3)-* packages conflict on each other for core packages [14:22] jamespage: figured i'd mention that in case you can think of an issue with that. for example with some packages like horizon dashboards files conflict in /usr/share === tikund is now known as sikun [17:28] rbasak: if you are back, and/or feeling able, could you please import into git-ubuntu pmdk and ndctl? They were accepted into cosmic (\o/) === Guest20971 is now known as maco === led_ir23 is now known as led_ir22