Switch to DuckDuckGo Search
   May 3, 2020  
< | 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 | >

Toggle Join/Part | bottom
[00:04:14] *** idodeclare <idodeclare!~textual@cpe-76-185-177-63.satx.res.rr.com> has quit IRC (Ping timeout: 240 seconds)
[00:04:16] <neirac> toastersonerson I missed a char, yes, now trying to run the container failed with this https://pastebin.com/mfkv1CC5
[00:06:40] <toastersonerson> ah thats a race condition. just start it again
[00:10:36] *** phyre <phyre!~phyre___@78.30.23.107> has quit IRC (Remote host closed the connection)
[00:13:58] <neirac> toastersonerson I tried 10 times, I always endup with this https://pastebin.com/jYJ7vseq
[00:14:22] <toastersonerson> can you try with zoneadm directly?
[00:14:34] <toastersonerson> it's still truncating the command output
[00:17:42] <neirac> toastersonerson root/sbin/init: No such file or directory
[00:17:55] *** idodeclare <idodeclare!~textual@cpe-76-185-177-63.satx.res.rr.com> has joined #illumos
[00:17:57] <toastersonerson> ???
[00:18:12] <toastersonerson> did it's rootfs not properly mount?
[00:18:16] <neirac> could not stat /podhost/pods/aef0c363-66cc-403e-83eb-4fa3531a1e3b/root/sbin/init: No such file or directory
[00:18:37] <toastersonerson> ls /podhost/pods/aef0c363-66cc-403e-83eb-4fa3531a1e3b/
[00:18:59] <toastersonerson> something is wrong with the zfs layout.
[00:19:13] <toastersonerson> probably a bug during dreation of the tarfile
[00:19:26] <toastersonerson> a root dir got introduced accidentally
[00:19:43] <toastersonerson> but i need a dir layout to see whats wrong
[00:19:46] <neirac> on that dir I see dev pod.json and root directories
[00:20:04] <toastersonerson> in the directory directly?
[00:20:46] <toastersonerson> cat you paste that dir and the root dir contents?
[00:21:38] <neirac> toastersonerson here it is https://pastebin.com/nfvag80x
[00:22:43] <toastersonerson> that checks out. is there nothing under sbin?
[00:24:43] <neirac> toastersonerson 0 files 0 directories under /podhost/pods/aef0c363-66cc-403e-83eb-4fa3531a1e3b/root/sbin
[00:25:20] <neirac> the zone was created using a smartos image in first place and using the illumos brand
[00:25:21] <toastersonerson> is that the pkgsrc brand?
[00:25:30] <toastersonerson> oh
[00:25:40] <toastersonerson> yep. lofs mounts are missing :)
[00:25:40] <neirac> toastersonerson is the illumos brand
[00:25:53] <toastersonerson> are you on smartos?
[00:25:57] <neirac> in omnios
[00:26:23] <toastersonerson> you will need to switch the brand to the appropriate one with zonecfg
[00:27:01] <toastersonerson> note to self. import brand if present in zone metadata
[00:28:43] <neirac> toastersonerson the zone has brand pod the one that was created from the image created by imageadm
[00:29:05] <toastersonerson> yes you will need to change it to the one that the imported zone is
[00:29:29] <toastersonerson> pod is my brand which is equivalent to the alien brand of tribblix
[00:29:33] <neirac> toastersonerson oh ok, seems like a small fix on imageadm right?
[00:29:54] <neirac> toastersonerson oh the alien brand I just knew about it a couple of days ago
[00:29:56] <toastersonerson> Yeah I need to save more zone metadata when importing
[00:30:13] <toastersonerson> you can change via zonecfg
[00:30:18] <neirac> toastersonerson let me try looking at the code to add that
[00:30:32] <neirac> toastersonerson I'll do it and try again
[00:30:35] <toastersonerson> you will also want to save it in the annotations of the image
[00:30:46] <toastersonerson> so that it gets saved properly
[00:31:19] <toastersonerson> and then the container code needs to respect that annotation
[00:35:46] <neirac> how do I change hte brand using zonecfg?
[00:36:26] <LeftWing> You may be able to "set brand"?
[00:37:18] <neirac> Zone aef0c363-66cc-403e-83eb-4fa3531a1e3b already installed; set brand not allowed.
[00:37:56] <toastersonerson> in the xml then
[00:38:22] <toastersonerson> but i think openindiana allows one to do so
[00:38:33] <toastersonerson> I remember doing so at least. but some time ago
[00:38:39] <jbk> you may need to set it prior to installing the zone
[00:44:16] <Smithx10> toastersonerson: So I imagine the Kubelet and Nomad Agents that support OCI just fork out to run podman? https://www.nomadproject.io/docs/drivers/external/podman/
[00:44:25] *** sjorge <sjorge!~sjorge@unaffiliated/sjorge> has quit IRC (Ping timeout: 265 seconds)
[00:44:41] *** sjorge <sjorge!~sjorge@unaffiliated/sjorge> has joined #illumos
[00:45:04] <toastersonerson> Smithx10 (IRC): kubelet wants a CRI implementing daemon
[00:45:19] <toastersonerson> for nomad the driver can use the API's directly without forking
[00:48:13] <Smithx10> Seems k8s hasn't stopped growing hahaha
[00:48:15] <Smithx10> https://kubernetes.io/docs/setup/production-environment/container-runtimes/
[00:49:08] <toastersonerson> yes those are implementations of a CRI
[00:49:38] <neirac> toastersonerson I took the uuid from podadm list, then modified the brand in xml, but still give the same error running the container
[00:49:47] <Smithx10> So you'll have like a CRI-O, or a CRI-Illumos?
[00:50:16] <toastersonerson> yeah looks like I need to check out brand support
[00:50:25] <toastersonerson> CRI-illumos yes
[00:53:07] <neirac> let me try a pkgsrc zone
[00:53:28] <toastersonerson> that will faill guaranteed.
[00:53:58] <neirac> toastersonerson, that will have brand pkgsrc and it will not use hte joyent image
[00:54:19] <toastersonerson> sparse based brands will not have their lofs mounts set up _)
[00:54:22] <toastersonerson> :)
[00:54:43] <neirac> toastersonerson oh ok, that's on go-zone ?
[00:55:13] <toastersonerson> no thats the same problem of the brand
[00:55:25] <toastersonerson> it is hardcoded as brand=pod ATM
[00:55:42] <neirac> toastersonerson also I removed zones using zoneadm now podadm list shows a lot of warnings
[00:56:13] <toastersonerson> yes there are json files you will need to remove
[00:56:20] <toastersonerson> under /podhost
[00:56:24] <toastersonerson> * under /podhost/pods
[01:06:15] <kamil_warszawa> what is the purpose of solaris contracts?
[01:06:20] <kamil_warszawa> and why zones need them
[01:07:26] <toastersonerson> neirac (IRC): I am heading to sleep now, but matrix stays online. I am adressable. Good night
[01:07:58] <neirac> toastersonerson good night, thanks for your help
[01:08:57] <toastersonerson> No problem. It's good to have first users :)
[01:09:25] <tomww> contracts are useful for servce management, you can find some SMF services herding a zoo of processes
[01:09:54] <kamil_warszawa> do zones depend on them?
[01:10:18] <tomww> especially if you can't identify one specific PID as a master. E.g. if just one proecess must exist and lived with that contract id, then this SMF service is in shape
[01:11:58] <tomww> there are several places where a process carries IDs in the process context. one is "projects", other is "contracts" and if I remember correctly for zones is another ID. that way the kernel can see which process blongs to a zone and what is allowed to a process to access and see
[01:14:06] <kamil_warszawa> are there equivalents in linux or bsds?
[01:15:17] <tomww> equivalent conects in linux or bsd?
[01:15:40] <kamil_warszawa> to contracts
[01:15:43] <tomww> only bsd jails may share ideas with zones.
[01:16:08] <kamil_warszawa> are contracts needed in zones?
[01:16:13] <kamil_warszawa> are they essential
[01:16:16] <tomww> I didn't look closer at linux
[01:16:28] <tomww> contracts have nothing essential to zones.
[01:18:03] <tomww> why should they? there is already the zones-id carried along with every process runinning in that zone. so the kernel can always filter on that ID to grant access to a ressource or deny. it is _not_ that simple if you look into the code, but you see the concept.
[01:18:30] <LeftWing> Contracts are a different type of grouping of processes
[01:18:51] <tomww> yes, the purpose of contracts is different
[01:18:53] <LeftWing> You could have zones without contracts, and contracts without zones -- they're effectively orthogonal
[01:19:18] <kamil_warszawa> ah, good!
[01:19:52] <kamil_warszawa> utilities: zlogin, zoneadm, zoneadmd, zonename, zonecfg, zonestat, zoneinfo -- is this a complete list of cmds for zones?
[01:20:21] <tomww> all those are used outside of zone but for managing zones
[01:21:36] <LeftWing> I don't think we have a command called "zoneinfo"
[01:23:25] <kamil_warszawa> ah, I mixed it with TZ dir
[01:23:47] <kamil_warszawa> have I missed any other zone tool?
[01:25:35] *** Teknix <Teknix!~pds@69.41.134.110> has quit IRC (Ping timeout: 258 seconds)
[01:27:07] *** Teknix <Teknix!~pds@69.41.134.110> has joined #illumos
[01:52:37] *** ldepandis <ldepandis!~ldepandis@unaffiliated/ldepandis> has quit IRC (Quit: My MacBook has gone to sleep. ZZZzzz…)
[02:35:54] <Smithx10> toastersonerson: does podman support events ?
[02:36:52] <Smithx10> I’m just thinking of the challenges that vmadm has with scaling without vminfod
[03:37:34] <kamil_warszawa> are zones TM by oracle?
[03:52:11] *** nbhauke_ <nbhauke_!~hauke@55d46542.access.ecotel.net> has joined #illumos
[03:53:24] *** nbhauke <nbhauke!~hauke@55d41425.access.ecotel.net> has quit IRC (Ping timeout: 265 seconds)
[03:53:24] *** nbhauke_ is now known as nbhauke
[04:24:59] *** TheGreekOwl <TheGreekOwl!~Arkomeda@2a02:587:4874:d500:69f1:5548:4980:af2d> has quit IRC (Ping timeout: 265 seconds)
[04:31:17] *** Teknix <Teknix!~pds@69.41.134.110> has quit IRC (Ping timeout: 260 seconds)
[04:33:35] *** Teknix <Teknix!~pds@69.41.134.110> has joined #illumos
[04:53:24] *** jcea <jcea!~Thunderbi@2001:bc8:2ecd:caed:7670:6e00:7670:6e00> has quit IRC (Quit: jcea)
[05:37:11] *** kamil_warszawa <kamil_warszawa!~kamil@188.146.230.17.nat.umts.dynamic.t-mobile.pl> has quit IRC (Remote host closed the connection)
[07:27:33] *** BOKALDO <BOKALDO!~BOKALDO@91.105.124.91> has joined #illumos
[07:31:24] *** idodeclare <idodeclare!~textual@cpe-76-185-177-63.satx.res.rr.com> has quit IRC (Ping timeout: 272 seconds)
[07:33:26] *** idodeclare <idodeclare!~textual@cpe-76-185-177-63.satx.res.rr.com> has joined #illumos
[09:44:14] *** tsoome__ <tsoome__!~tsoome@91.209.240.229> has quit IRC (Read error: Connection reset by peer)
[09:45:02] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has quit IRC (Read error: Connection reset by peer)
[09:49:44] *** tsoome_ <tsoome_!~tsoome@91.209.240.229> has joined #illumos
[09:57:59] *** nbhauke <nbhauke!~hauke@55d46542.access.ecotel.net> has quit IRC (Quit: nbhauke)
[09:58:11] *** nbhauke <nbhauke!~hauke@55d46542.access.ecotel.net> has joined #illumos
[10:17:01] *** clapont <clapont!~clapont@unaffiliated/clapont> has quit IRC (Ping timeout: 264 seconds)
[10:18:20] *** clapont <clapont!~clapont@unaffiliated/clapont> has joined #illumos
[10:24:44] *** nbhauke <nbhauke!~hauke@55d46542.access.ecotel.net> has quit IRC (Quit: nbhauke)
[10:33:54] *** vgusev_ <vgusev_!~vgusev@188.187.60.230> has joined #illumos
[10:36:58] *** vgusev <vgusev!~vgusev@188.187.60.230> has quit IRC (Ping timeout: 272 seconds)
[10:50:46] *** nbhauke <nbhauke!~hauke@55d46542.access.ecotel.net> has joined #illumos
[11:03:48] *** nbhauke <nbhauke!~hauke@55d46542.access.ecotel.net> has quit IRC (Quit: nbhauke)
[11:03:52] *** phyre <phyre!~phyre___@78.30.23.107> has joined #illumos
[11:14:03] *** KeiraT <KeiraT!~k4ra@gateway/tor-sasl/k4ra> has quit IRC (Ping timeout: 240 seconds)
[11:16:55] *** KeiraT <KeiraT!~k4ra@gateway/tor-sasl/k4ra> has joined #illumos
[12:20:43] *** ptribble <ptribble!~ptribble@cpc92716-cmbg20-2-0-cust138.5-4.cable.virginm.net> has joined #illumos
[12:23:09] *** leoric <leoric!~leoric@46.147.100.3> has joined #illumos
[12:28:34] <neuroserve> don't think that "zones" is trademarkable - maybe "solaris zones" is
[12:28:34] *** wacki <wacki!~wacki@i577B843D.versanet.de> has joined #illumos
[12:31:49] <igork> zones - can be interpretative as container
[13:04:15] <toastersonerson> Smithx10 (IRC): No i was not aware of events.
[13:04:27] <toastersonerson> Iare there any docs and sample what you mean?
[13:09:57] *** nbhauke <nbhauke!~hauke@55d46542.access.ecotel.net> has joined #illumos
[13:14:40] *** TheGreekOwl <TheGreekOwl!~Arkomeda@2a02:587:4874:d500:69f1:5548:4980:af2d> has joined #illumos
[13:17:01] *** nbhauke <nbhauke!~hauke@55d46542.access.ecotel.net> has quit IRC (Ping timeout: 264 seconds)
[13:20:06] *** nbhauke <nbhauke!~hauke@55d46542.access.ecotel.net> has joined #illumos
[14:19:37] *** nbhauke <nbhauke!~hauke@55d46542.access.ecotel.net> has quit IRC (Quit: nbhauke)
[14:20:43] <toastersonerson> Smithx10 (IRC): Ah i see yes I was already having some performance problems. A daemon for lookups is an interesting idea.
[14:22:33] *** wiedi <wiedi!~wiedi@ip5b4096a6.dynamic.kabel-deutschland.de> has quit IRC (Quit: ^C)
[14:25:04] *** phyre <phyre!~phyre___@78.30.23.107> has quit IRC (Remote host closed the connection)
[14:28:26] *** jcea <jcea!~Thunderbi@2001:bc8:2ecd:caed:7670:6e00:7670:6e00> has joined #illumos
[14:30:41] *** cypa <cypa!~cypam]_@5.79.187.109> has joined #illumos
[14:42:20] <toastersonerson> Although one can make a state file update without a daemon
[14:42:42] <toastersonerson> I will need to find out how to integrate with native tooling though
[15:26:10] *** tsoome <tsoome!~tsoome@148-52-235-80.sta.estpak.ee> has joined #illumos
[16:10:56] *** liv3010m <liv3010m!~liv3010m@77-72-245-190.fibertel.com.ar> has quit IRC (Ping timeout: 256 seconds)
[16:11:39] *** liv3010m <liv3010m!~liv3010m@77-72-245-190.fibertel.com.ar> has joined #illumos
[16:40:47] *** jellydonut <jellydonut!~quassel@s91904423.blix.com> has quit IRC (Quit: jellydonut)
[16:42:46] *** jellydonut <jellydonut!~quassel@s91904423.blix.com> has joined #illumos
[16:44:33] *** baojg <baojg!~baojg@2604:a880:0:1010::16:1> has joined #illumos
[16:48:56] *** ldepandis <ldepandis!~ldepandis@unaffiliated/ldepandis> has joined #illumos
[17:07:28] <Smithx10> toastersonerson: Would still be nice to go find out where you spend time... Not sure if zones / zfs / network can come up any faster
[17:10:17] *** ChrisBF <ChrisBF!56802735@host86-128-39-53.range86-128.btcentralplus.com> has joined #illumos
[17:19:44] *** wiedi <wiedi!~wiedi@ip5b4096a6.dynamic.kabel-deutschland.de> has joined #illumos
[17:31:34] <toastersonerson> Smithx10 (IRC): You mean where the image spends time?
[17:48:27] <Smithx10> Well create, delete, etc
[17:51:35] <toastersonerson> ah yeah. some integration with opentracing will be required once I am at that point
[17:51:47] <toastersonerson> Thsts the golang native way to see these things
[17:52:21] <toastersonerson> It's an extension of the prometheus monitoring stack
[17:52:31] <toastersonerson> Have alook at Jaeger for sample apps
[17:59:14] *** ChrisBF <ChrisBF!56802735@host86-128-39-53.range86-128.btcentralplus.com> has quit IRC (Ping timeout: 245 seconds)
[18:12:17] *** idodeclare <idodeclare!~textual@cpe-76-185-177-63.satx.res.rr.com> has quit IRC (Ping timeout: 265 seconds)
[18:15:14] *** ptribble <ptribble!~ptribble@cpc92716-cmbg20-2-0-cust138.5-4.cable.virginm.net> has quit IRC (Quit: Leaving)
[18:22:22] *** tsoome_ <tsoome_!~tsoome@91.209.240.229> has quit IRC (Read error: Connection reset by peer)
[18:23:19] *** tsoome_ <tsoome_!~tsoome@91.209.240.229> has joined #illumos
[18:26:17] <Smithx10> Yeah I used jaeger it was pretty cool
[18:26:46] <Smithx10> But I mean in the OS, not sure if anyone has ever optimized or looked at making them go a bit faster
[18:26:53] <toastersonerson> Thinking of improvements?
[18:27:04] <toastersonerson> dtrace :)
[18:27:16] <Smithx10> Yeah, just not sure if anyone has looked
[18:27:36] <toastersonerson> Circonous has IIRC. YOu might want to ask THeo Schlossnagle
[18:30:18] *** idodeclare <idodeclare!~textual@cpe-76-185-177-63.satx.res.rr.com> has joined #illumos
[18:30:22] *** baojg <baojg!~baojg@2604:a880:0:1010::16:1> has quit IRC (Remote host closed the connection)
[19:03:24] *** andy_js <andy_js!~andy@90.218.209.121> has joined #illumos
[19:13:51] *** phyre <phyre!~phyre___@78.30.23.107> has joined #illumos
[19:15:45] *** phyre <phyre!~phyre___@78.30.23.107> has quit IRC (Remote host closed the connection)
[19:41:05] *** cypa <cypa!~cypam]_@5.79.187.109> has quit IRC (Read error: Connection reset by peer)
[19:41:06] *** cypa_ <cypa_!~cypam]_@5.79.187.109> has joined #illumos
[19:55:34] *** clapont <clapont!~clapont@unaffiliated/clapont> has quit IRC (Ping timeout: 240 seconds)
[19:58:14] *** clapont <clapont!~clapont@unaffiliated/clapont> has joined #illumos
[19:58:37] *** tomww <tomww!~tom@unaffiliated/tomww> has quit IRC (Remote host closed the connection)
[19:58:48] *** leah2 <leah2!~leah@vuxu.org> has quit IRC (Ping timeout: 265 seconds)
[20:16:26] *** tomww <tomww!~tom@unaffiliated/tomww> has joined #illumos
[20:19:51] *** leah2_ <leah2_!~leah@vuxu.org> has joined #illumos
[20:23:20] *** leah2_ is now known as leah2
[20:23:20] *** leah2 <leah2!~leah@vuxu.org> has quit IRC (Excess Flood)
[20:25:31] *** leah2_ <leah2_!~leah@vuxu.org> has joined #illumos
[20:27:15] *** leah2_ is now known as leah2
[20:29:29] *** ngchk1 <ngchk1!~ngchk1@b2b-92-50-91-166.unitymedia.biz> has joined #illumos
[20:48:27] *** leoric <leoric!~leoric@46.147.100.3> has quit IRC (Quit: Konversation terminated!)
[20:50:50] *** leoric <leoric!~leoric@46.147.100.3> has joined #illumos
[21:19:23] *** leoric <leoric!~leoric@46.147.100.3> has quit IRC (Quit: Konversation terminated!)
[21:21:46] <neirac> would be to much work to run linux binaries in the gz ?, just thought of running steam in illumos would be nice.
[21:28:05] <tomww> you need the interpretation layer to get linux kernel interface. that is doen with the LX-zone-brand
[21:30:22] <neirac> I think in FreeBSD they have hardcoded paths were the linux libraries should be loaded
[21:44:44] <tomww> you can't run linux binaries natively on Illumos kernel
[21:45:02] <tomww> the point is: runnin glinux binaries is only possible inside a Zone with the lx-brand
[21:47:43] <tomww> there is a layer that translates calls coming from the linux binary into a Illumos kernel call. there is documentation on the net how that works for the lx-brand and solaris-10 brand.
[21:48:16] <neirac> tomww thanks
[21:59:51] *** BOKALDO <BOKALDO!~BOKALDO@91.105.124.91> has quit IRC (Quit: Leaving)
[22:15:22] <KungFuJesus> tsoome_: let me know if you bake a new USB image to try out for SPARC, Oracle killed T2 support in 11.4
[22:18:54] <andyf> I'm trying to debug a bit of a mess with pkcs11...
[22:19:20] <andyf> or, at least the nss plugin to libkmf
[22:19:42] <andyf> It ends up including <pkcs11t.h> and <pkcs11pub.h>
[22:19:52] <andyf> but the former comes from gate, and the latter from NSS
[22:20:14] <andyf> NSS has pkcs11t.h too..
[22:20:36] <andyf> and a new type has been added in NSS 3.52 (CK_GENERATOR_FUNCTION)
[22:20:57] <andyf> the type is referenced in NSS' pkcs11pub.h but it not present in gate's pkcs11t.h
[22:21:05] <andyf> Is mess the right word?
[22:21:39] *** nbhauke <nbhauke!~hauke@55d46542.access.ecotel.net> has joined #illumos
[22:22:20] <tomww> ;)
[22:22:37] <andyf> Adding the new type gate's pkcs11t.h gets things compiling again
[22:25:38] *** wacki <wacki!~wacki@i577B843D.versanet.de> has left #illumos
[22:29:42] *** phyre <phyre!~phyre___@78.30.23.107> has joined #illumos
[22:42:40] <jbk> it sounds like they've already mandated v3.0 support
[23:02:52] *** nbhauke <nbhauke!~hauke@55d46542.access.ecotel.net> has quit IRC (Quit: nbhauke)
[23:05:11] <andyf> It's definitely one of the new 3.0 interfaces
[23:08:26] <wilbury> why is is_pseudo_device() function not published in sys/ includes?
[23:08:37] <wilbury> (although more drivers are using it)
[23:11:17] *** andy_js <andy_js!~andy@90.218.209.121> has quit IRC (Ping timeout: 260 seconds)
[23:11:34] *** andy_js <andy_js!~andy@90.218.209.121> has joined #illumos
[23:16:59] *** andy_js <andy_js!~andy@90.218.209.121> has quit IRC (Read error: Connection reset by peer)
[23:18:09] *** andy_js <andy_js!~andy@90.218.209.121> has joined #illumos
[23:31:46] *** andy_js <andy_js!~andy@90.218.209.121> has quit IRC (Quit: andy_js)
[23:36:44] <jbk> i'd like to get things updated to 3.0, but pkcs11_softtoken really needs a lot of work first
[23:36:51] <jbk> so many bugs
[23:45:01] <LeftWing> wilbury: Unclear. It should be in a header, even if it's one we don't ship outside the gate.
[23:45:35] <LeftWing> Perhaps uts/common/sys/ddi_implfuncs.h or somethiung
[23:58:29] <wilbury> LeftWing: yes, it should. i noticed in ramdisk.c (which i take, among others, as a reference for my kernel driver coding exercise)
[23:59:16] <wilbury> and also in vds.c (ldom virtual disk)
top

   May 3, 2020  
< | 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 | >