January 14, 2019 
< | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31 | >
January 14, 2019 [00:11:03] <jbk> so?[00:11:53] <jbk> the json is the same format for all VMs[00:13:18] <jbk> i have a router zone that's setup to use 256MB of ram[00:13:22] <jbk> as well as a dns server[00:13:26] <jbk> because that's all they need[00:13:59] <rsully> jbk native zones or? curious what software you're using for routing and dns :)[00:14:30] <rsully> but yeah I have some that are 512MB ram, MB is really most intuitive for everything[00:19:29] <jbk> yeah native zones[00:20:00] <jbk> but the json for a native zone isn't radically different than an HVM one -- yeah some fields change, but it's more or less the same[00:32:08] <jrg> hm.....[00:34:44] <jrg> so can i use the joyent windows server 2012 iso for 2016?[00:35:08] *** ingenthr <ingenthr!~ingenthr@47.150.244.29> has joined #smartos[00:44:03] *** Gathis <Gathis!~TheBlack@unaffiliated/gathis> has quit IRC (Remote host closed the connection)[01:00:19] *** ingenthr <ingenthr!~ingenthr@47.150.244.29> has quit IRC (Quit: Leaving.)[01:03:53] *** axonpoet <axonpoet!~axonpoet@fsf/member/axonpoet> has joined #smartos[01:14:37] *** axonpoet <axonpoet!~axonpoet@fsf/member/axonpoet> has quit IRC (Ping timeout: 246 seconds)[01:15:10] <jrg> seems so[01:15:29] <jrg> well. that’s neat.[01:22:31] *** andy_js <andy_js!~andy@94.3.60.133> has quit IRC (Quit: andy_js)[02:08:58] *** jellydonut <jellydonut!~jelly@77.18.55.110.tmi.telenormobil.no> has quit IRC (Ping timeout: 272 seconds)[02:23:47] *** v_a_b <v_a_b!~volker@p4FECC531.dip0.t-ipconnect.de> has quit IRC (Ping timeout: 240 seconds)[02:48:24] <Smithx10> awk indexes at 1 ? WHAT?!$$![02:48:27] <Smithx10> Crime.....[03:18:51] *** ingenthr <ingenthr!~ingenthr@47.150.244.29> has joined #smartos[03:19:57] *** ingenthr <ingenthr!~ingenthr@47.150.244.29> has quit IRC (Client Quit)[03:21:16] *** ingenthr <ingenthr!~ingenthr@47.150.244.29> has joined #smartos[03:22:19] *** ingenthr <ingenthr!~ingenthr@47.150.244.29> has quit IRC (Client Quit)[03:23:35] *** ingenthr <ingenthr!~ingenthr@47.150.244.29> has joined #smartos[03:24:44] *** ingenthr <ingenthr!~ingenthr@47.150.244.29> has quit IRC (Client Quit)[03:26:05] *** ingenthr <ingenthr!~ingenthr@47.150.244.29> has joined #smartos[03:45:58] <jrg> well.. managed to get w10 pro working on smartos[03:46:37] <jrg> seems pretty solid too other than the 2 sockets 2 cpus heh[03:46:45] <jrg> not sure if i can even bump that up[03:47:31] <Smithx10> mgerdts: https://github.com/joyent/mi-centos-bhyve-7/issues/2[03:47:44] <Smithx10> Latest CentOS image defaults to 8.8.88[03:48:03] <Smithx10> Different from the Ubuntu Images.[04:11:22] *** axonpoet <axonpoet!~axonpoet@fsf/member/axonpoet> has joined #smartos[04:16:07] *** axonpoet <axonpoet!~axonpoet@fsf/member/axonpoet> has quit IRC (Ping timeout: 258 seconds)[04:51:26] *** jcea <jcea!~Thunderbi@2001:41d0:1:8a82:7670:6e00:7670:6e00> has quit IRC (Quit: jcea)[05:02:56] *** black_pete <black_pete!~black_pet@gateway.peterguy.com> has joined #smartos[06:20:29] *** rsully <rsully!~rsully@unaffiliated/rsully> has quit IRC (Quit: rsully)[06:35:32] *** ingenthr <ingenthr!~ingenthr@47.150.244.29> has quit IRC (Ping timeout: 268 seconds)[06:44:31] *** saltsa <saltsa!~joonas@dsl-hkibng42-58c3f6-87.dhcp.inet.fi> has quit IRC (Quit: leaving)[07:21:50] *** _Tenchi_ <_Tenchi_!~phil@207-255-80-203-dhcp.aoo.pa.atlanticbb.net> has quit IRC (Ping timeout: 272 seconds)[07:25:12] *** _Tenchi_ <_Tenchi_!~phil@207-255-80-203-dhcp.aoo.pa.atlanticbb.net> has joined #smartos[07:29:28] *** _Tenchi_ <_Tenchi_!~phil@207-255-80-203-dhcp.aoo.pa.atlanticbb.net> has quit IRC (Ping timeout: 246 seconds)[07:35:40] *** ingenthr <ingenthr!~ingenthr@47.150.244.29> has joined #smartos[07:54:20] *** v_a_b <v_a_b!~volker@p57A27EC0.dip0.t-ipconnect.de> has joined #smartos[07:58:34] *** ingenthr <ingenthr!~ingenthr@47.150.244.29> has quit IRC (Quit: Leaving.)[08:11:40] *** neuroserve <neuroserve!~toens@195.71.113.124> has joined #smartos[08:18:11] *** _Tenchi_ <_Tenchi_!~phil@207-255-80-203-dhcp.aoo.pa.atlanticbb.net> has joined #smartos[08:28:09] *** hotbox <hotbox!~hotbox@2001:41d0:fe8f:b70a:20d:b9ff:fe47:7c05> has joined #smartos[08:35:21] *** black_pete <black_pete!~black_pet@gateway.peterguy.com> has quit IRC (Quit: black_pete)[08:36:57] *** mno-hime <mno-hime!~mno-hime@94.142.238.232> has quit IRC (Remote host closed the connection)[08:37:30] *** JeanParpaillon <JeanParpaillon!~jean@176.164.111.172> has joined #smartos[08:37:46] *** mno-hime <mno-hime!~mno-hime@94.142.238.232> has joined #smartos[08:42:02] *** mno-hime <mno-hime!~mno-hime@94.142.238.232> has quit IRC (Client Quit)[09:04:20] *** olsner <olsner!~salparot@c83-249-185-75.bredband.comhem.se> has quit IRC (Ping timeout: 250 seconds)[09:40:14] *** JeanParpaillon <JeanParpaillon!~jean@176.164.111.172> has quit IRC (Ping timeout: 246 seconds)[10:07:34] *** bens1 <bens1!~bens1@161.122.2.81.in-addr.arpa> has joined #smartos[10:16:19] *** Gathis <Gathis!~TheBlack@unaffiliated/gathis> has joined #smartos[10:35:25] *** man_u <man_u!~manu@manu2.gandi.net> has joined #smartos[10:45:51] *** jellydonut <jellydonut!~jelly@77.16.68.144.tmi.telenormobil.no> has joined #smartos[10:45:57] <kayront> is there any effort at all at the moment in getting smartos/illumos to run on other platforms such as arm(64) and openpower?[10:46:31] <jperkin> kayront: mostly stalled at this point, help would be very welcome![11:12:26] *** andy_js <andy_js!~andy@94.3.60.133> has joined #smartos[11:16:50] <kayront> is triton also a gui, like project-fifo?[11:17:46] <kayront> speaking of project-fifo, is it dead? channel is +m and licenser idle for 3 months, heh[11:18:27] <jperkin> triton comes with an admin interface which lets you manage things via a web interface, the bit that isn't included is a way for customers and non-admins to do the same with their VMs[11:20:58] <kayront> i see. and about project-fifo, do you know why the lack of news etc?[11:21:32] <jperkin> I've never used it, sorry[11:21:59] <kayront> np[11:23:15] <kayront> one last question, are you able to recommend well-supported, low-power consumption, quiet hardware for smartos in 2019? i'm looking at 16-32G RAM and enough room for 2-4 disks[11:23:33] <kayront> for a homelab[11:23:51] <kayront> bonus points if it serves coffee[11:23:55] <jperkin> I know a bunch of people have had good success with intel nucs, obviously would use external disks, but..[11:24:11] <jperkin> personally I'm still running on an ancient HP N36L so not a good person to ask ;)[11:24:27] <jperkin> (though does fulfil most of the brief)[11:24:44] <kayront> yeah i've had the nuc recommended here before as well, fishing for other options/opinions[11:24:59] <kayront> going to pull the trigger on this soon™[11:25:26] <jperkin> if there was a newer microserver at a reasonable price I'd upgrade, but available options don't seem worth it, especially with crazy DRAM prices[11:30:09] <kayront> amd processors won't work for kvm right?[11:30:20] <kayront> this was the case some years ago iirc[11:30:21] <jperkin> correct, at least not without external patches[11:36:47] <kayront> http://caiustheory.com/compiling-smartos-for-amd-processors/[11:36:53] <kayront> would this still work? it's a few years old[11:37:12] <kayront> any particular reason why the official image doesn't have these patches baked in?[11:37:52] <jperkin> person to ask about all that stuff is LeftWing, he'll be awake in the next few hours (US west coast)[11:39:37] <kayront> cheers[12:34:13] *** strobelight <strobelight!~quassel@c-69-180-7-202.hsd1.ga.comcast.net> has quit IRC (Ping timeout: 245 seconds)[13:00:24] <neuroserve> kayront: https://docs.danubecloud.org/user-guide/index.html is based on smartos, too[13:59:34] *** jellydonut <jellydonut!~jelly@77.16.68.144.tmi.telenormobil.no> has quit IRC (Read error: Connection reset by peer)[14:08:32] *** strobelight <strobelight!quassel@nat/cisco/x-whwhfwrganjlzwom> has joined #smartos[14:11:45] *** andy_js_ <andy_js_!~andy@94.3.60.133> has joined #smartos[14:17:32] *** Gathis <Gathis!~TheBlack@unaffiliated/gathis> has quit IRC (Remote host closed the connection)[14:17:38] <mamalos> OK, here's the question. I'm playing with containerpilot, because I like its idea, and[14:17:41] <mamalos> have the following config (for testing): https://pastebin.com/LhhKnvsV[14:18:34] <mamalos> with this config, happy-app-nginx job's service is never registered in consul, whereas happy-app-nginx-public does. This doesn't change if exec becomes "service nginx start"[14:18:58] <mamalos> I am not sure how to troubleshoot it, hence my question.[14:19:18] <mamalos> and one more question: Are there any plans of integrating nomad in Triton (through the API I mean)?[14:19:31] <mamalos> I'm running it within containers using the raw_exec and exec drivers, but it would be interesting if it could be integrated with cloud-api (just like docker and kubernetes)[14:19:46] <mattronix> hi[14:20:00] <mattronix> How does one monitor for disk failure in a Trition SDC env?[14:20:05] <mattronix> any Amon examples?[14:50:59] <neuroserve> mattronix: I guess most of the Triton-Users deploy a local monitor for that[14:57:14] *** jellydonut <jellydonut!~jelly@77.16.68.144.tmi.telenormobil.no> has joined #smartos[15:15:27] *** Kurlon_ <Kurlon_!~Kurlon@98.13.72.207> has quit IRC (Read error: Connection reset by peer)[15:15:33] *** Kurlon <Kurlon!~Kurlon@98.13.72.207> has joined #smartos[15:21:43] *** Kurlon <Kurlon!~Kurlon@98.13.72.207> has quit IRC (Ping timeout: 245 seconds)[15:22:22] *** andy_js_ <andy_js_!~andy@94.3.60.133> has quit IRC (Quit: andy_js_)[15:32:10] *** jcea <jcea!~Thunderbi@2001:41d0:1:8a82:7670:6e00:7670:6e00> has joined #smartos[15:36:49] *** andy_js_ <andy_js_!~andy@94.3.60.133> has joined #smartos[15:52:23] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has joined #smartos[15:52:45] *** jcea <jcea!~Thunderbi@2001:41d0:1:8a82:7670:6e00:7670:6e00> has quit IRC (Remote host closed the connection)[15:53:04] *** jcea <jcea!~Thunderbi@2001:41d0:1:8a82:7670:6e00:7670:6e00> has joined #smartos[16:10:46] *** polishdub <polishdub!~polishdub@207.86.38.254> has joined #smartos[16:27:44] *** ingenthr <ingenthr!~ingenthr@47.150.244.29> has joined #smartos[16:38:37] <mgerdts> @Smithx10 thanks for reporting. For the sake of clarity as time goes on, can you include the uuid of the problematic image in the bug report?[16:38:57] <Smithx10> Sure, its in the output of the run command[16:39:03] <Smithx10> itt create *[16:39:07] <Smithx10> ill update it tho[16:40:23] <Smithx10> 1f11188a-c71c-11e8-83d4-370e2a698b16[16:43:52] *** andy_js_ <andy_js_!~andy@94.3.60.133> has quit IRC (Quit: andy_js_)[16:48:06] *** ingenthr <ingenthr!~ingenthr@47.150.244.29> has quit IRC (Quit: Leaving.)[16:48:14] <mgerdts> ahh, wasn't familiar with your itt command.[16:49:49] <jrg> hm. seems like windows defender doesn't do well with rdp on my win10 vm running on smartos[17:00:06] <jrg> seems like the vdisk goes super slow too[17:05:26] <mgerdts> jrg are you running a platform image that is 20181206 or later?[17:05:35] *** ingenthr <ingenthr!~ingenthr@47.150.244.29> has joined #smartos[17:11:32] <Smithx10> sorry mgerdts I alias that stuff :([17:11:41] <Smithx10> its just insecure triton, sinc we dont have certs setup yet at work[17:11:57] <mgerdts> what, no desire to hack json every time?[17:13:31] <Smithx10> i think its just triton -i *[17:16:21] <muchomas> when you get around to the certs, i wholeheartedly endorse triton-dehydrated[17:16:48] *** mmalecki <mmalecki!~maciej@homie.mmalecki.com> has quit IRC (Remote host closed the connection)[17:17:03] <jrg> platform image?[17:17:12] <jrg> oh. yes. i just updated yesterday[17:17:39] <jrg> and created the windows vm yesterday as well.. it runs at least heh. but the whole point was to use rsat for my samba AD[17:20:03] <jrg> it's ok. i guess i just installed an older version of rsat.. i have the dns app now using the ws2016 ver[17:20:20] *** neuroserve <neuroserve!~toens@195.71.113.124> has quit IRC (Ping timeout: 272 seconds)[17:20:46] <jrg> i wonder how much ram i can reduce this to. windows 10 will run on 2GB?[17:21:18] <jrg> i'm actually a bit confused about running out of disk space on the host. i should have had more than it allowed me to give to the windows VM[17:23:13] <jrg> zones/c367c0da-a863-6836-bf36-f3f57abb6184-disk0 volsize 512G local[17:23:22] <jrg> but zfs shows something a bit odd[17:23:44] <jrg> zones/c367c0da-a863-6836-bf36-f3f57abb6184-disk0 762G 574G 235G -[17:24:23] <jrg> i don't understand how a zvol that has a size of 512GB is using 762GB[17:24:40] <mgerdts> I don't know what the RAM requirements for windows 10 is.[17:25:18] <mgerdts> can you run `zfs get all zones/c367c0da-a863-6836-bf36-f3f57abb6184-disk0` and paste it to a gist, pastebin, or similar?[17:25:48] <mgerdts> oh, you are using kvm. there's no recent improvements to disk or network performance for that.[17:26:18] <mgerdts> also zfs list -r -t all zones/c367c0da-a863-6836-bf36-f3f57abb6184-disk0[17:29:51] <Smithx10> mgerdts: I think LX has an issue alsow ith centos[17:30:01] <Smithx10> Where if /etc/resolv.conf exists it wont populate or delete it[17:30:05] <Smithx10> and its the opposite for ubuntu hahahaha[17:30:11] <Smithx10> ./strange days.[17:31:16] <jrg> https://justpaste.it/4dbsb[17:32:13] <jrg> zones/c367c0da-a863-6836-bf36-f3f57abb6184-disk0 762G 574G 235G -[17:32:16] <jrg> zones/c367c0da-a863-6836-bf36-f3f57abb6184-disk0@2019-01-12_1330 2.25G - 235G -[17:33:41] <jrg> i don't see why the used is like 700GB[17:33:53] <jrg> 762[17:35:52] <jrg> i know at one point i resized the zvol and changed its size in the VM but the extra 200+GB does not make sense[17:36:00] <jrg> can i change the used flag?[17:41:30] <Latrina> guys good evening[17:41:46] <Latrina> can I create a vnic that is not binded to any physical nic?[17:42:10] <jrg> oh. it was the refreservation[17:42:14] *** greenline <greenline!~user@unaffiliated/greenline> has joined #smartos[17:42:42] <Latrina> what I need to do is basiscally forward the OpenVPN traffic from one lx zone to this dummy vnic[17:42:59] <Latrina> and make it act as gw for other zones to use the OpenVPN connection[17:43:23] <wiedi> Latrina: "nictagadm add -l" creates etherstubs[17:43:50] <Latrina> oh, so what I need is an etherstubs? I will read about it. Thanks[17:44:10] <jrg> is refreservation sometrhing i really needed to keep on? i'm guessing that is a just in case moment if your disks actually run out of space for the dataset?[17:45:33] *** greenline <greenline!~user@unaffiliated/greenline> has quit IRC (Client Quit)[17:47:36] *** aweinstock <aweinstock!~aweinstoc@cpe-67-248-65-250.nycap.res.rr.com> has joined #smartos[17:48:14] *** greenline <greenline!~user@unaffiliated/greenline> has joined #smartos[17:52:14] *** black_pete <black_pete!~black_pet@gateway.peterguy.com> has joined #smartos[18:00:02] *** jcea <jcea!~Thunderbi@2001:41d0:1:8a82:7670:6e00:7670:6e00> has quit IRC (Remote host closed the connection)[18:00:16] *** jcea <jcea!~Thunderbi@2001:41d0:1:8a82:7670:6e00:7670:6e00> has joined #smartos[18:01:50] <jrg> yeah. seems like disk performance in kvm / win10 is seriously slow. good thing i'm just using it for administrative purposes which doesn't really need much disk speed. it seems to function well tho so there's that[18:03:42] <jesse_> jrg, kvm does all IO as sync IO[18:04:05] <jesse_> you need a fast slog device to make it faster[18:04:34] <jrg> jesse_: ah. ok. that explains the slower performance in the linux VMs then. i used all kvm.[18:04:35] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has quit IRC (Ping timeout: 246 seconds)[18:04:42] <jrg> maybe later I’ll try creating a bhyve VM and see how that works out[18:04:58] <jesse_> for linux non-kernel stuff I'd use an lx zone[18:05:21] <jrg> i thought about that but my linux installs need smb mounting which seems a bit more difficult to do on the host level with smartos[18:05:39] <jesse_> I think there was something about not being able to run both kvm and bhyve on the same host, too[18:05:46] <jrg> ouch. really?[18:06:09] <jrg> i was about to ask about that.. maybe swapping from kvm to bhyve if that was possible[18:06:26] <jesse_> may have been just some glitch when bhyve was still (more) in the works[18:06:28] <jrg> i can possibly migrate everything to my freenas box with zfs send then create bhyve VMs .. then migrate it back[18:06:43] <jrg> all I’d have to do is migrate the disks[18:07:11] <jrg> they’re just zvols. i had no problem migrating a bhyve zvol to smartos and running it with kvm[18:07:13] <jesse_> there are probably some posts about converting from kvm to bhyve on the internet[18:07:32] <jrg> that would be nice to find. I’d probably try that out with the win10 VM first[18:07:47] <jesse_> I haven't needed bhyve/kvm in ages[18:07:52] <rmustacc> You should be able tor un both bhyve and kvm at the same time now, IIRC.[18:07:54] <jesse_> I should probably try bhyve[18:07:58] *** sjorge <sjorge!~sjorge@unaffiliated/sjorge> has quit IRC (Quit: 410 Gone)[18:07:58] <pmooney> If you're on a recent PI, KVM and bhyve can coexist on the same system[18:08:32] <jesse_> nice[18:08:33] <jrg> pmooney: oh? so I’d have to migrate the zvols... wipe all the VMs... then recreate them... then move them back?[18:08:36] *** sjorge <sjorge!~sjorge@unaffiliated/sjorge> has joined #smartos[18:08:42] <jrg> or is that not an option either?[18:09:15] <greenline> svc:/smartdc/agent/ca/cainstsvc:default is failed in my headnode gz, and I don't particularly care about it. is there a way to permanently disable? I know how to add services via /opt/custom - add a service-disabling service?[18:09:16] <jrg> i wouldn’t mind taking the lx route but I’d have to figure out setting up smb mounting on the smartos host into the lx zones[18:09:34] <jrg> i think I’d wind up with awkward permission issues tho[18:09:54] <pmooney> jrg: you might be able to accomplish it by creating a diskless bhyve instance and renaming your zvols over[18:10:40] <jrg> pmooney: that would be a difficult thing to do with 5 kvms.. i wouldn’t mind doing it if i didn’t have a couple important ones running. i noticed the network performance of kvm isn’t that great as well[18:11:00] <pmooney> Yeah, I didn't say it'd be easy, haha[18:11:20] <pmooney> And depending on the guest, some OSes might be agitated about the PCI layout changing[18:12:15] <jrg> i just stuck with kvm because it’s a bit more along than bhyve[18:12:20] <jrg> bhyve has been flakey in my experience[18:12:32] <pmooney> how so?[18:13:04] <jrg> getting win10 to install was an undertaking... i also had issues when i updated to a newer ver of fbsd[18:13:06] <jrg> where VMs just outright stopped running[18:13:18] <jrg> although I’ve never tried it on smartos[18:13:25] <jrg> nor have i tried it recently[18:14:08] <jrg> my personal favorite was when i updated fbsd and a zimbra vm stopped working for some reason and rolling back didn’t help... i wound up using zfs send to the smartos box to get it back up and running in kvm[18:14:41] <jrg> talk about high strung that day[18:14:46] <steven43126> Upgraded platform image and hitting https://smartos.org/bugview/TRITON-766, whats the correct workaround for this for now ?[18:15:19] <steven43126> https://smartos.org/bugview/TRITON-884 also looks relevant in that it looks like it will be ultimatley getting removed[18:15:33] <jbk> steven43126: that's correct -- you can just ignore the error[18:16:20] <steven43126> jbk: there is mention that some sdcadm commands might fail? it also trips our monitoring because it flags in fmadm faulty[18:16:45] <steven43126> is the workaround at the bottom of https://smartos.org/bugview/TRITON-884 safe ? ie removing the service ?[18:16:49] <jrg> at some point when i have time i might throw another 1u in and try to migrate my current VMs to it and use triton vs smartos[18:17:01] <jrg> and bhyve[18:17:45] <jbk> trent is probably the person that'd know better but that looks like it just deletes the smf instance, which should be ok[18:18:08] <mgerdts> jrg I think the used is so high because the snapshot of it is referencing most of the "referenced" blocks. The refreservation is ensuring that the pool has enough space that you can never do a write in the guest that turns into a ENOSPC (in host) / EIO (in guest).[18:18:48] <mgerdts> removing the refreservation is fine until your disk space gets tight.[18:19:17] <jrg> mgerdts: that’s what i did for now. i have half the pool free so i’m not too worried about it but I’ll keep an eye on it[18:19:44] <jrg> although i’m curious as to which snapshot[18:20:04] <mgerdts> zones/c367c0da-a863-6836-bf36-f3f57abb6184-disk0@2019-01-12_1330[18:20:08] <mgerdts> from your zfs list output[18:20:29] <jrg> ah. you’re right. i created that before i updated the guest OS[18:20:39] <mgerdts> which is a great idea, btw. :)[18:20:56] <jrg> well i didn’t want to have a “moment”[18:21:05] <greenline> oh, my question above was answered by the command at the bottom of https://smartos.org/bugview/TRITON-884[18:21:06] <jrg> i usually snapshot before major updates[18:22:06] <jrg> mgerdts: but yeah. i didn’t even notice that the referenced size was 235G in the snapshot... should have done the math there[18:23:48] <Latrina> wiedi: I can't understand what it wants[18:23:52] <Latrina> [root@homelab ~]# nictagadm -v add -l lol[18:23:53] <mgerdts> figuring out how zfs uses space and reports it is quite the undertaking.[18:23:53] <Latrina> + Using config file: /usbkey/config[18:23:55] <Latrina> Error: nic tag name is invalid[18:24:20] <Latrina> do I need to edit /usbkey/config and specify the etherstub in the first ?[18:24:25] <jrg> mgerdts: yes. it can be. i was lost as to why i couldn’t create a larger Win10 VM because i was “out of space” when i should have had like 1.3TB free[18:24:52] <jrg> and why the centos VM was reporting that extra 200+GB[18:25:19] <mgerdts> When you create the next one, it will take a huge refreservation too. That space will be unavailable to the extant VMs, which could lead them to ENOSPC/EIO much sooner.[18:25:58] *** bens1 <bens1!~bens1@161.122.2.81.in-addr.arpa> has quit IRC (Quit: bens1)[18:26:22] <jrg> I’m starting to wish i had used platters instead of SSDs for more space jic[18:27:36] <jrg> i suppose i can one by one them with platters and let it auto expand later. i figured I’d get better performance off 4 SSDs vs platters but that doesn’t seem to be the case due to the kvm bottleneck.[18:28:03] <jrg> which i suppose is the reason for the push to move to bhyve[18:29:05] <jrg> there are quite a few things i had known before deciding to swap to smartos ... otherwise it has been rock solid though[18:34:14] <Smithx10> Does LX have a python bug?[18:34:57] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has joined #smartos[18:35:31] <Latrina> wiedi: added ovpnint1_nic=08:00:27:c1:50:36 in /usbkey/config and it looks like it worked[18:40:13] <Latrina> nope, if I do that it is seen as a normal interface...[18:44:51] <Latrina> now it got added, however I am cant seem to figure how to assign it a MAC[18:44:53] <Latrina> nictagadm add -l ovpnint1 08:00:27:c1:50:36[18:44:55] <Latrina> Error: create local tag, but mac address specified[18:45:02] <Latrina> I wonder if this is how it is supposed to work...[18:45:33] *** theandro1edan <theandro1edan!~pfranz@c-174-59-156-209.hsd1.pa.comcast.net> has quit IRC (Read error: Connection reset by peer)[18:46:15] <Latrina> from the man[18:46:17] <Latrina> ic tags are used in SmartOS to refer to a physical nic without needing[18:46:19] <Latrina> its underlying MAC address or interface name. B[18:46:35] <Latrina> right, sorry for the flood[18:47:41] <Smithx10> Looks like there a docker image that is failling on a certain python call :([18:54:39] *** theandromedan <theandromedan!~pfranz@c-174-59-156-209.hsd1.pa.comcast.net> has joined #smartos[18:58:21] *** mno-hime <mno-hime!~mno-hime@94.142.238.232> has joined #smartos[18:58:58] *** man_u <man_u!~manu@manu2.gandi.net> has quit IRC (Ping timeout: 246 seconds)[19:18:15] <Smithx10> iperf3 is showing me 0.00000000000 from ubuntu BHYVE over the fabric...... .?[19:22:29] <Smithx10> Its only when we go over the overlay[19:22:40] <Smithx10> https://paste.ec/paste/Od0S-2Di#0SPOUd5QWxC2eHK4m4T-R7fVO8+1CzT22/xzGxaslYr[19:22:56] <Smithx10> rzezeski: have you seen this before?[19:23:32] <Smithx10> Im running on joyent_20190103T010943Z[19:24:15] <bahamat> Are you using udp?[19:24:49] <Smithx10> the default iperf3 command[19:25:09] <Smithx10> non fabric works[19:25:44] <Smithx10> ubuntu@brads-ubuntu:~$ iperf3 -c 192.168.128.23[19:25:54] <Smithx10> and ubuntu@b4daf08f-f853-4919-d79a-ddc68a89fe0c:~$ iperf3 -s 192.168.128.23[19:32:39] <Smithx10> I believe that’s tcp[19:33:09] <rzezeski> default is TCP, and it looks like data is getting dropped (what is the underlying NIC?)[19:34:26] <Smithx10> x520[19:35:46] *** rsully <rsully!~rsully@unaffiliated/rsully> has joined #smartos[19:36:49] <Smithx10> rzezeski: from top of rack network https://paste.ec/paste/Z5XgtnWa#Y7Cwz+J+n3JurbAqC5UAEUyVQFxoWgby0tVPXCsjma3[19:36:53] <Smithx10> same servers[19:37:16] <Smithx10> those are ubuntu bhyve machines..... going to try smartos machines now[19:47:41] <Smithx10> When doing it on SmartOS in smartos base64 I got 1 [ 4] 0.00-1.00 sec 194 KBytes 1.59 Mbits/sec . and then all 0's[19:51:25] <rzezeski> I think there's a chance this could be OS-7501[19:51:26] <jinni> Sorry, OS-7501 is not public.[19:52:10] <Smithx10> :( ./womp ./womp not public[19:52:26] <Smithx10> What is the ping command to test max mtu ?[19:52:35] <Smithx10> isnt it just ping addr size?[20:08:24] <bahamat> Smithx10: -D for Don't fragment[20:09:05] <wiedi> Latrina: i think nictag names should end with a number[20:21:50] <LeftWing> Latrina: When creating an etherstub (-l), you don't give it a MAC address. The manual page sort of says this.[20:22:15] <LeftWing> In addition, in general NIC tags shouldn't end with a number, either.[20:23:09] <Latrina> LeftWing: yeah I read the man ultimately[20:23:26] <Smithx10> bahamat: ping -D 192.168.128.24 9500 ?[20:23:29] <Smithx10> it still pings :([20:23:30] <LeftWing> Though I guess for an etherstub it's OK for them to end in a number[20:23:36] <Smithx10> it should crash and burn right?[20:23:45] <Latrina> as far as the name is concerned, I get all kind of bad names... even when using a dot or underscore[20:23:45] <LeftWing> Or, even, they _must_ end in a number. I should have coffee.[20:24:14] <Latrina> I guess it would be nice to have some kind of --force-badnames feature lol[20:24:17] * LeftWing was thinking about Triton nictags[20:26:34] <Smithx10> ping -D -s 192.168.128.24 9000[20:26:35] <Smithx10> w00t[20:26:55] <Latrina> I will be replacing ESXi with SmartOS on my APU2C4 once I am done with the relocation[20:27:14] <Latrina> my OVPN server and main virtswitch runs on it[20:28:38] <bahamat> Smithx10: Yeah, it looks like the size is only set if -s is used.[20:29:30] <bahamat> LeftWing: Is is normal that it ignores stuff unless you use -s?[20:31:18] <bahamat> I guess the man page intends to say that, but it's very unclear that that's the case.[20:31:23] <LeftWing> Yeah I think if you don't pass -s as the first option it's operating in the first mode listed in the synopsis in the manual page[20:31:44] <LeftWing> Well, the "-s" option is not in [optional] brackets[20:31:48] <LeftWing> in the second synopsis[20:32:04] <bahamat> Yeah, it mentions the data size in the same paragraph as the -s, but it doesn't explicitly say that data size is only changable with -s.[20:32:31] <LeftWing> Without -s you're in the first synopsis entry, though, which has only two possible arguments[20:32:47] <LeftWing> We could certainly add a sentence to DESCRIPTION to try and clear that up, though[20:32:53] <bahamat> Yeah.[20:37:01] <Smithx10> looks like MTU is wrong[20:37:04] <Smithx10> [root@f8-f2-1e-3b-09-c4 (us-east-1) ~]# ping -D -s 10.79.4.2 1535[20:37:04] <Smithx10> PING 10.79.4.2: 1535 data bytes[20:37:21] <Smithx10> doesnt work but 1534 does.... these are supposed to be 9000[20:45:11] <Smithx10> to confirm that should work on mtu 9000 right?[20:51:08] *** axonpoet <axonpoet!~axonpoet@fsf/member/axonpoet> has joined #smartos[20:51:49] <bahamat> Smithx10: Are both the network and nic_tag set to 9000?[20:52:05] <Smithx10> Yes[20:52:18] <bahamat> Can you show me?[20:52:48] <bahamat> This is for the underlay itself.[20:52:54] <bahamat> Not the fabric networks.[21:00:08] *** polishdub <polishdub!~polishdub@207.86.38.254> has quit IRC (Quit: leaving)[21:16:08] <Latrina> LeftWing: can a MTU of 9000 be used for a etherstub?[21:18:17] <LeftWing> Latrina: Yes, I believe so[21:18:37] <Latrina> thanks, I will try that[21:18:38] <Latrina> :)[21:20:03] <Latrina> Error: Cannot specify MTU for local nic tag[21:20:28] <Latrina> dladm set-linkprop -p mtu=9000 happyether works though[21:20:52] <Latrina> not sure if the mtu has really been changed at this point[21:21:10] <LeftWing> Actually, doing "dladm show-linkprop -p mtu $etherstub" suggests it's already 9000 for my etherstubs[21:21:13] <LeftWing> Maybe we make them all 9000[21:22:19] <Latrina> oh[21:22:22] <Latrina> nice[21:25:09] <Latrina> but for instance, now whatever name I put in nictagadm add -l throws out Error: nic tag name is invalid[21:25:18] <Latrina> and is exactly the issue I was having before[21:25:33] <Latrina> you suggest a name and sure it will fail[21:25:48] <Latrina> I am not sure if this is a bug at this point[21:26:19] <LeftWing> What name are you trying to create[21:26:29] <Latrina> I tried happyether[21:26:32] <LeftWing> I think I might have been extremely confusing earlier -- etherstubs _do_ need to end in a number[21:26:34] <LeftWing> so happyether0[21:26:40] <Latrina> ovpn, etherbuds, lol nothing works[21:27:00] <LeftWing> Sorry about that[21:27:01] <Latrina> so happyether0 should work?[21:27:07] <LeftWing> I think so yes[21:27:15] <Latrina> sorry? no need to say sorry man[21:27:23] <Latrina> worked[21:27:25] <Latrina> ..[21:27:29] <Latrina> so it needs to end with a number?[21:27:36] <LeftWing> For etherstubs (-l) yes[21:27:52] <Latrina> okay. I will not it down. thanks[21:43:34] *** Gathis <Gathis!~TheBlack@unaffiliated/gathis> has joined #smartos[21:56:58] *** theandromedan <theandromedan!~pfranz@c-174-59-156-209.hsd1.pa.comcast.net> has quit IRC (Ping timeout: 250 seconds)[21:59:33] *** theandromedan <theandromedan!~pfranz@c-174-59-156-209.hsd1.pa.comcast.net> has joined #smartos[22:02:53] *** ingenthr <ingenthr!~ingenthr@47.150.244.29> has quit IRC (Quit: Leaving.)[22:04:54] *** morgib_ <morgib_!uid342182@gateway/web/irccloud.com/x-klbeszibadhibcjr> has joined #smartos[22:04:56] *** mno-hime <mno-hime!~mno-hime@94.142.238.232> has quit IRC (Quit: Leaving)[22:07:23] *** hotbox_ <hotbox_!~hotbox@2001:41d0:fe8f:b70a:20d:b9ff:fe47:7c05> has joined #smartos[22:11:39] *** Teknix <Teknix!~pds@2601:801:101:f429:f1ce:b6f7:4dc2:223f> has quit IRC (Ping timeout: 268 seconds)[22:11:41] *** hotbox <hotbox!~hotbox@2001:41d0:fe8f:b70a:20d:b9ff:fe47:7c05> has quit IRC (Ping timeout: 268 seconds)[22:11:42] *** gemelen <gemelen!~gemelen@zooey.gemelen.net> has quit IRC (Ping timeout: 268 seconds)[22:11:45] *** Dominic <Dominic!~dominic@chaldene.m0dlx.com> has quit IRC (Ping timeout: 268 seconds)[22:11:47] *** morgib <morgib!uid342182@gateway/web/irccloud.com/x-yuuouonsykjyzzqd> has quit IRC (Ping timeout: 268 seconds)[22:11:51] *** morgib_ is now known as morgib[22:12:12] *** Teknix <Teknix!~pds@2601:801:101:f429:55af:6e03:287:af3c> has joined #smartos[22:13:48] *** mpana <mpana!~mpana@193.226.149.49> has quit IRC (Ping timeout: 246 seconds)[22:14:08] *** tuxillo <tuxillo!~antonioh@89.128.97.235> has quit IRC (Ping timeout: 246 seconds)[22:23:47] <Smithx10> bahamat: I just confirmed they didn’t set jumbo frames on the switch[22:23:59] <bahamat> That'll do it.[22:24:09] <Smithx10> Network teams are awesome[22:24:54] <jbk> just remember, the network is never at fault![22:24:55] <jbk> :)[22:25:18] <jbk> at a previous job, I had to argue with a network person on how ARP worked[22:25:43] *** mmalecki <mmalecki!~maciej@144.76.114.210> has joined #smartos[22:26:05] <jbk> 'i can ping the gateway, i'm sending packets to the gateway, they never arrive at the destination, the destination can also reach it's default gateway, they're on separate vlans, how would static ARP fix this?'[22:27:07] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has quit IRC (Ping timeout: 240 seconds)[22:41:27] *** ingenthr <ingenthr!~ingenthr@47.150.244.29> has joined #smartos[22:42:16] *** ingenthr <ingenthr!~ingenthr@47.150.244.29> has quit IRC (Client Quit)[22:42:29] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has joined #smartos[22:42:40] *** ingenthr <ingenthr!~ingenthr@47.150.244.29> has joined #smartos[22:43:28] *** ingenthr <ingenthr!~ingenthr@47.150.244.29> has quit IRC (Client Quit)[22:44:51] *** ingenthr <ingenthr!~ingenthr@47.150.244.29> has joined #smartos[22:45:54] *** ingenthr <ingenthr!~ingenthr@47.150.244.29> has quit IRC (Client Quit)[22:47:46] *** ingenthr <ingenthr!~ingenthr@47.150.244.29> has joined #smartos[22:49:28] <Smithx10> lololol[22:50:15] *** ingenthr <ingenthr!~ingenthr@47.150.244.29> has quit IRC (Client Quit)[22:51:33] <jbk> finally about about 20-30 mins of arguing with the guy, i just asked him 'ok, what entry should i put in there'[22:51:41] <jbk> and then the lightbulb went off[22:52:00] *** ingenthr <ingenthr!~ingenthr@47.150.244.29> has joined #smartos[22:53:42] <Smithx10> lolol[22:53:46] <Smithx10> you have to play inception with them[23:05:58] *** Kurlon_ <Kurlon_!~Kurlon@bidd-pub-04.gwi.net> has joined #smartos[23:07:31] *** rsully <rsully!~rsully@unaffiliated/rsully> has quit IRC (Quit: rsully)[23:08:59] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has quit IRC (Ping timeout: 268 seconds)[23:10:24] *** Kurlon_ <Kurlon_!~Kurlon@bidd-pub-04.gwi.net> has quit IRC (Ping timeout: 258 seconds)[23:20:23] <Latrina> https://gist.github.com/psammarco/9c3f1c39e3ae3fc74c103d026b4f9d0d#file-create-etherstub-vnic-add-it-to-zone-smartos[23:20:40] <Latrina> I put together a mini howoto of my findings today[23:46:36] *** BrownBear <BrownBear!~BrownBear@107.175.247.247> has quit IRC (Ping timeout: 268 seconds)[23:47:09] *** Kurlon <Kurlon!~Kurlon@98.13.72.207> has joined #smartos[23:57:47] <Latrina> guys I have a question[23:57:56] <Latrina> how do I change the kernel in a lx zone?[23:58:23] <Latrina> should I proceed as you normally would on a linux distro or is there something specific smartos related?[23:58:34] <bahamat> You can't really.[23:58:35] <Smithx10> This is starting to burn me with people moving their docker images to 18.04[23:58:38] <Smithx10> https://github.com/joyent/sdc-docker/issues/136[23:58:45] <Latrina> damn[23:58:56] <Smithx10> :([23:59:02] <bahamat> You can change the reported kernel version using vmadm, but that doesn't affect the behavior of the actual running kernel at all.[23:59:31] <Latrina> tun is not implemented in the kernel that comes with alpinelinux in this lx zone[23:59:53] <Latrina> damn, I guess I need to go KVM