Switch to DuckDuckGo Search
   July 23, 2017
< | 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

bottom
[00:01:08] <bennabiy> I just rebooted a kvm and it is giving me ECONNREFUSED when I try to do vmadm info UUID vnc
[00:01:18] <bennabiy> how can I get access to it?
[00:01:25] <nbjoerg> check the port it is listen on
[00:01:31] <nbjoerg> unless you explicitly set it, will be random
[00:01:35] <bennabiy> I know
[00:01:41] <bennabiy> this is from the GZ
[00:02:05] <bennabiy> # vmadm info 2ea31676-faed-e8d5-cfb6-9526b0735ae6 vnc
[00:02:06] <bennabiy> Unable to get VM info for VM 2ea31676-faed-e8d5-cfb6-9526b0735ae6: connect ECONNREFUSED
[00:03:18] <Smithx10> bennabiy: tried killing it?
[00:03:31] <bennabiy> I stopped it, and this is while starting it
[00:07:48] <bennabiy> one thing I did was try what LeftWing and others suggested of moving away from vnc to vmadm console, but I do not think I did it right. I will probably have to boot it to a live disk, and revert my grub changes
[00:08:27] <nbjoerg> if you can set up a serial console, it is generally much easier to use
[00:10:01] <bennabiy> I did something along the lines of the changes made by this script https://github.com/joyent/mi-debian-8/search?utf8=✓&q=ttyS0&type=
[00:11:13] <bennabiy> and it worked until I rebooted (forced to reboot ) and now I cannot get a connection to it and vmadm cannot communicate with it nor zlogin
[00:11:46] *** andy_js <andy_js!~andy@94.3.225.134> has quit IRC (Quit: andy_js)
[00:12:10] *** nikolam <nikolam!~nikolam@unaffiliated/nikolam> has quit IRC (Remote host closed the connection)
[00:16:46] <bennabiy> what is the vmadm command to boot to "cd" temporarily for a kvm?
[00:22:16] *** bahamat <bahamat!~bahamat@2001:470:f0f4:a0a:8000:ee1a:6aeb:f931> has joined #smartos
[00:25:01] <nbjoerg> boot order=d or so
[00:36:58] *** Smithx10 <Smithx10!~smithx10@106.19.62.94.rev.vodafone.pt> has quit IRC (Remote host closed the connection)
[00:37:25] *** Smithx10 <Smithx10!~smithx10@106.19.62.94.rev.vodafone.pt> has joined #smartos
[00:48:40] *** jhendricks <jhendricks!~jhendrick@c-73-162-7-40.hsd1.ca.comcast.net> has joined #smartos
[00:51:32] *** jhendricks1 <jhendricks1!~jhendrick@8.12.47.4> has joined #smartos
[00:53:17] *** jhendricks <jhendricks!~jhendrick@c-73-162-7-40.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 260 seconds)
[00:54:35] *** deep-book-gk_ <deep-book-gk_!~1wm_su@5.62.43.15> has joined #smartos
[00:57:32] *** deep-book-gk_ <deep-book-gk_!~1wm_su@5.62.43.15> has left #smartos
[01:07:47] <LeftWing> bennabiy: Which of those things did you do?  The "/etc/inittab" thing, or adding "console=ttyS0,115200n8" to GRUB_CMDLINE_LINUX_DEFAULT in your grub configuration and running update-grub?
[01:07:49] *** Gathis <Gathis!~TheBlack@unaffiliated/gathis> has quit IRC (Remote host closed the connection)
[01:14:03] <bennabiy> I think I did both
[01:14:21] <bennabiy> It is an encrypted root
[01:15:12] <bennabiy> If you have suggestions of how to connect, I am all ears
[01:16:11] <bennabiy> I either need to revert the changes I made to grub, or access it how it it
[01:16:14] <bennabiy> it is*
[01:22:08] <bennabiy> LeftWing: ^
[01:22:41] <bennabiy> I know for sure I did the grub configuration
[01:32:06] <bennabiy> I am trying to boot to a live cd to get access to the drive to check its configuration
[01:32:46] <bennabiy> I am getting ECONNREFUSED with that as well
[01:34:41] *** jocthbr <jocthbr!uid240867@gateway/web/irccloud.com/x-xqdnhrkrosstzdqd> has joined #smartos
[01:35:00] <bennabiy> actually it is not an encrypted root
[01:36:11] *** jocthbr <jocthbr!uid240867@gateway/web/irccloud.com/x-xqdnhrkrosstzdqd> has quit IRC (Client Quit)
[01:41:18] <bennabiy> If I need to roll back to a previous snapshot of a disk in a KVM, what is the best way to preserve the current state? zfs send to file?
[01:43:05] *** bahamat <bahamat!~bahamat@2001:470:f0f4:a0a:8000:ee1a:6aeb:f931> has quit IRC (Quit: Leaving.)
[01:45:56] *** almostworking <almostworking!~almostwor@unaffiliated/almostworking> has quit IRC (Ping timeout: 240 seconds)
[01:51:36] *** jcea <jcea!~Thunderbi@ks27448.kimsufi.com> has quit IRC (Remote host closed the connection)
[01:55:57] <LeftWing> bennabiy: What is the state of the zone at this time?  (How does it appear in "vmadm list")
[01:59:03] <LeftWing> Also, for what it's worth, I don't think you want to do the inittab thing
[01:59:28] <LeftWing> (I think that's something that's being done there specifically in the context of the install ISO?)
[01:59:49] <LeftWing> You really just want to do whatever is standard for your Linux distribution to change the system console to ttyS0
[02:00:01] <LeftWing> There shouldn't really be anything SmartOS-specific about that.
[02:01:43] <bennabiy> LeftWing: running at this moment
[02:02:19] <bennabiy> LeftWing: I know I did the grub change, not sure if I tried the inittab
[02:02:30] <LeftWing> OK
[02:02:38] <LeftWing> And you can't access it over the network?
[02:02:41] <bennabiy> I am making a backup of my current disk0 and going to try to rollback
[02:02:53] <bennabiy> no because it hangs waiting for encryption key to be input
[02:02:55] <bennabiy> no ssh yet
[02:02:56] <LeftWing> Well, let's try and figure out why you cannot get the "vmadm info" bit to work first
[02:03:19] <bennabiy> ok
[02:03:33] <bennabiy> I have a backup .zfs file of the current disk0
[02:03:37] <bennabiy> just incase
[02:03:40] <LeftWing> OK
[02:04:35] <bennabiy> # vmadm info 2ea31676-faed-e8d5-cfb6-9526b0735ae6
[02:04:35] <bennabiy> Unable to get VM info for VM 2ea31676-faed-e8d5-cfb6-9526b0735ae6: connect ECONNREFUSED
[02:04:47] <LeftWing> truss -f -t connect -v connect -a vmadm info 2ea31676-faed-e8d5-cfb6-9526b0735ae6
[02:05:22] <bennabiy> EINPROGRESS
[02:05:29] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[02:05:31] <bennabiy> for the connect
[02:06:52] <LeftWing> to  "AF_INET  name = 127.0.0.1  port = 8080" ?
[02:07:19] <bennabiy> https://pastebin.mozilla.org/9027810
[02:07:46] <LeftWing> Alright
[02:08:08] <LeftWing> In another window, can you run:     tail -f $(svcs -L vmadmd) | bunyan -o short
[02:08:19] <LeftWing> Then try the "vmadm info" again without "truss" ?
[02:09:38] <bennabiy> ok
[02:09:45] <bennabiy> same error
[02:10:01] <bennabiy> nothing appeared on the other window
[02:10:35] <bennabiy> LeftWing: ^
[02:11:09] <LeftWing> Nothing?!
[02:11:12] <LeftWing> Not, e.g., 00:10:57.295Z DEBUG vmadmd: url /vm/1ab3c66f-5f27-e770-b62d-c7c21f7a8361/info?types=all
[02:11:15] <bennabiy> Nothing!
[02:11:27] <LeftWing> That's not ideal.  Is vmadmd actually running?  "svcs -xv"
[02:12:28] <bennabiy> https://pastebin.mozilla.org/9027811
[02:12:39] <bennabiy> not running!
[02:12:51] <LeftWing> Uh oh
[02:12:54] <LeftWing> What's in those log files?
[02:12:58] <bennabiy> about to look
[02:13:02] <LeftWing> e.g., bunyan $(svcs -L vmadmd)
[02:13:08] <LeftWing> and bunyan $(svcs -L metadata)
[02:13:56] <bennabiy> what is the escape for bunyan?
[02:14:15] <LeftWing> It's just running "less", so you use the regular less "q" to get out.
[02:14:46] <LeftWing> Well, it's running $PAGER, which is very often "less".
[02:15:12] <bennabiy> not responding to CTRL-C
[02:15:20] <bennabiy> and q just echos to my screen
[02:15:31] <LeftWing> What exactly did you run?
[02:16:02] <bennabiy> I accidentally ran bunyan $(svcs -L vmadm) instead of vmadmd
[02:16:08] <LeftWing> Oh
[02:16:10] <LeftWing> Try Ctrl+D
[02:16:18] <bennabiy> great
[02:16:19] <bennabiy> that worked
[02:17:14] <bennabiy> https://pastebin.mozilla.org/9027812
[02:17:21] <bennabiy> vmadmd empty
[02:18:43] <bennabiy> should I try starting vmadmd service?
[02:19:25] <bennabiy> wait, that wasn't full output
[02:20:59] <LeftWing> bennabiy: Perhaps try...    tail -n 100 $(svcs -L vmadmd) | bunyan
[02:21:13] <LeftWing> Just to get the last 100 lines
[02:21:20] <LeftWing> (before it presumably died for some reason)
[02:24:50] <bennabiy> vmadmd log empty
[02:25:38] <LeftWing> If you... ls $(svcs -L vmadmd)*
[02:25:46] <LeftWing> Are there any files other than "/var/svc/log/system-smartdc-vmadmd:default.log"
[02:26:14] <bennabiy> nope, nothing else
[02:26:27] <LeftWing> OK, well, let's try clearing the service!
[02:26:44] <LeftWing> In one window, I would:  tail -f $(svcs -L vmadmd) | bunyan -o short
[02:26:50] <LeftWing> Then in another, "svcadm clear vmadmd"
[02:26:54] <bennabiy> one thing that took place, the hard drive filled up and was blocking for a while
[02:26:55] <LeftWing> Let's see what happens!
[02:26:58] <LeftWing> Oh
[02:27:02] <bennabiy> I cleared the drive
[02:27:04] <LeftWing> Well, that might explain it.
[02:27:06] <bennabiy> yes
[02:27:10] <bennabiy> no logs could be written
[02:27:13] <LeftWing> Aye.
[02:27:15] <bennabiy> I just thought of that
[02:28:22] <LeftWing> You'll want to bring metadata back online too
[02:28:24] <bennabiy> that did something :)
[02:28:27] <LeftWing> Hooray!
[02:28:30] <LeftWing> Hopefully you can "vmadm info" now
[02:29:02] <bennabiy> yes
[02:29:27] <bennabiy> Is there a way to hardcode the same vnc port to a kvm
[02:29:30] <bennabiy> persistent?
[02:30:20] <AlainODea> PSA: Solaris ghc-7.10.2 bootstrap compiler used by PKGSRC segfaults on launching GHCi. I'm guessing it's either built wrong or built against Solaris 11. Joy. I'll have to start from GHC 7.6.3 after all...
[02:30:53] <LeftWing> bennabiy: Yes, if you take a look in the manual page I believe the VM JSON object has a "vnc_port" property.
[02:31:41] <bennabiy> ok, so I have my VM up in a liveCD.
[02:32:08] <bennabiy> checking /etc/default/grub for change
[02:33:42] <bennabiy> GRUB_CMDLINE_LINUX_DEFAULT="quiet divider=10 tsc=reliable console=ttyS0,115200n8 console=tty0"
[02:34:23] <bennabiy> LeftWing: I tried to set the vnc_port but after a reboot, it did not persist (reboot of the GZ that is)
[02:34:34] <LeftWing> How did you set it?
[02:34:45] *** dfs <dfs!~danilo@freebsd/user/dfs> has quit IRC (Quit: leaving)
[02:34:55] <bennabiy> my memory escapes me, but I thought with JSON through vmadm
[02:35:22] <LeftWing> You could try....   "vmadm update <uuid> vnc_port=5901"
[02:35:31] *** dfs <dfs!~danilo@freebsd/user/dfs> has joined #smartos
[02:36:04] <LeftWing> (for example)
[02:36:39] <bennabiy> yes, that changes it, which I think is what I did before...
[02:36:42] <bennabiy> I will see if it holds
[02:38:38] *** jocthbr <jocthbr!~quassel@138-122-44-143.host.cicloti.com.br> has joined #smartos
[02:41:14] <bennabiy> rebooting vm to see if console will work
[02:49:57] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Quit: Leaving.)
[02:59:37] *** v_a_b <v_a_b!~volker@p54BC7E4A.dip0.t-ipconnect.de> has quit IRC (Ping timeout: 255 seconds)
[03:03:42] *** v_a_b <v_a_b!~volker@p54BC755D.dip0.t-ipconnect.de> has joined #smartos
[03:09:46] *** strobelight <strobelight!~quassel@2001:420:c0c4:1005::7c> has quit IRC (Quit: gtg)
[03:20:00] *** ed209 <ed209!~ed209@165.225.128.67> has quit IRC (Remote host closed the connection)
[03:20:06] *** ed209 <ed209!~ed209@165.225.128.67> has joined #smartos
[03:34:33] *** jocthbr <jocthbr!~quassel@138-122-44-143.host.cicloti.com.br> has quit IRC (Remote host closed the connection)
[03:40:42] *** jocthbr <jocthbr!~quassel@138-122-44-143.host.cicloti.com.br> has joined #smartos
[04:16:12] *** Kurlon <Kurlon!~Kurlon@98.13.72.207> has quit IRC (Remote host closed the connection)
[04:16:47] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has joined #smartos
[04:20:25] *** tozhu <tozhu!~Thunderbi@101.207.125.86> has joined #smartos
[04:42:25] *** jhendricks1 <jhendricks1!~jhendrick@8.12.47.4> has quit IRC (Ping timeout: 240 seconds)
[04:42:49] *** Kurlon_ <Kurlon_!~Kurlon@98.13.72.207> has joined #smartos
[04:46:16] *** Kurlon <Kurlon!~Kurlon@bidd-pub-03.gwi.net> has quit IRC (Ping timeout: 255 seconds)
[05:14:55] <bennabiy> How is support for zfs encryption in smartos coming along? (data at rest)
[05:15:30] <bennabiy> LeftWing: thank you for the help earlier. It seems to be working again, although I am having to recover from the full disk issue
[05:26:01] *** mahtob <mahtob!~milisman@176.220.179.12> has joined #smartos
[05:26:17] *** mahtob <mahtob!~milisman@176.220.179.12> has left #smartos ("Leaving")
[05:44:34] *** bahamat <bahamat!~bahamat@2001:470:f0f4:a0a:8000:ee1a:6aeb:f931> has joined #smartos
[05:47:08] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[05:52:53] *** Smithx10 <Smithx10!~smithx10@106.19.62.94.rev.vodafone.pt> has quit IRC (Remote host closed the connection)
[05:56:16] *** flatline <flatline!~textual@cpe-172-248-198-0.socal.res.rr.com> has quit IRC (Ping timeout: 260 seconds)
[05:56:20] *** flatline_ <flatline_!~textual@cpe-172-248-198-0.socal.res.rr.com> has joined #smartos
[06:00:35] *** flatline <flatline!~textual@cpe-172-248-198-0.socal.res.rr.com> has joined #smartos
[06:01:52] *** flatline_ <flatline_!~textual@cpe-172-248-198-0.socal.res.rr.com> has quit IRC (Ping timeout: 255 seconds)
[06:20:13] *** jhendricks <jhendricks!~jhendrick@c-73-162-7-40.hsd1.ca.comcast.net> has joined #smartos
[06:21:47] *** jhendricks1 <jhendricks1!~jhendrick@8.12.47.4> has joined #smartos
[06:22:38] *** Smithx10 <Smithx10!~smithx10@106.19.62.94.rev.vodafone.pt> has joined #smartos
[06:24:27] *** jhendricks <jhendricks!~jhendrick@c-73-162-7-40.hsd1.ca.comcast.net> has quit IRC (Ping timeout: 240 seconds)
[06:26:57] *** Smithx10 <Smithx10!~smithx10@106.19.62.94.rev.vodafone.pt> has quit IRC (Ping timeout: 240 seconds)
[06:26:57] *** tozhu <tozhu!~Thunderbi@101.207.125.86> has quit IRC (Ping timeout: 260 seconds)
[06:27:29] <LeftWing> On the subject of ZFS encryption: I'm not sure!  I know people are working on it, and once it makes it into illumos-gate we'll have it in SmartOS shortly afterwards.  I don't know how far along that is, though.
[06:48:11] *** flatline_ <flatline_!~textual@cpe-172-248-198-0.socal.res.rr.com> has joined #smartos
[06:49:34] *** flatline <flatline!~textual@cpe-172-248-198-0.socal.res.rr.com> has quit IRC (Ping timeout: 255 seconds)
[06:54:59] *** tozhu <tozhu!~Thunderbi@171.221.44.177> has joined #smartos
[07:08:27] *** tozhu <tozhu!~Thunderbi@171.221.44.177> has quit IRC (Ping timeout: 240 seconds)
[07:24:35] *** flatline <flatline!~textual@cpe-172-248-198-0.socal.res.rr.com> has joined #smartos
[07:25:22] *** tozhu <tozhu!~Thunderbi@171.221.44.177> has joined #smartos
[07:25:52] *** flatline_ <flatline_!~textual@cpe-172-248-198-0.socal.res.rr.com> has quit IRC (Ping timeout: 260 seconds)
[07:47:45] *** MagnusDredd_ <MagnusDredd_!~MagnusDre@wsip-98-172-66-146.ph.ph.cox.net> has joined #smartos
[07:48:17] <MagnusDredd_> Anyone using LX Zones?
[07:48:28] <bahamat> Lots of people.
[07:49:15] <MagnusDredd_> I suppose I should have specified, “here”
[07:49:37] <bahamat> Not to be snarky, but again, "lots".
[07:51:12] <MagnusDredd_> What I’m interested in is whether the files within the zone are individually stored in ZFS snapshots
[07:51:51] <bahamat> LX zones have a zfs dataset in the same way that native zones do.
[07:53:00] <MagnusDredd_> …which would allow individual files to be pulled out of snapshots instead of the “all or nothing” issue with ZVOLs, correct
[07:53:19] <bahamat> Individual files can be pulled from snapshots.
[07:53:30] <bahamat> But with ZFS, datasets are at a directory level.
[07:53:40] <MagnusDredd_> Apologies for the stupid questions, I just hate to assume
[07:54:00] <bahamat> No worries.
[07:54:46] <bahamat> You can snapshot a dataset, i.e., a directory and all descendents. And you can restore individual files from the snapshot, or roll back the entire dataset.
[07:55:32] <MagnusDredd_> *nod*  been using ZFS for years now.  I just have no experience with LX Zones, or Zones for that matter
[07:55:48] <bahamat> FreeBSD user?
[07:56:56] <MagnusDredd_> I’ve not moved my main filevserver from OpenSolaris since it has just worked, but I’ve only used OpenSolaris for that...
[07:57:28] <bahamat> On SmartOS every zone has at least one ZFS dataset at the zone root (not to be confused with the zone filesystem root) that is managed by the system. Zones can also have an additional delegated dataset that can be managed by the zone tenant.
[07:57:29] <MagnusDredd_> For everything else it’s FreeBSD or Desktop OSX…
[07:58:13] <bahamat> If you've used OpenSolaris all this time but not used zones, you've been missing out ;-)
[07:58:27] <MagnusDredd_> The issue is that I don’t trust the state of support in Linux for ZFS
[07:58:47] <bahamat> But, zone management wasn't very easy or straightforward with OpenSolaris.
[07:59:04] <MagnusDredd_> my server is terribly underpowered
[07:59:10] <bahamat> Zone management is much more streamlined in SmartOS.
[07:59:37] <bahamat> If it's at least a 64-bit processor then SmartOS should work fine on it.
[07:59:44] <MagnusDredd_> It looks like it.  I’m not sure whether it will work nicely on any hardware that I have tho
[07:59:55] <MagnusDredd_> 1st generation Opteron
[08:00:14] <bahamat> There's a list of known working configurations here:
[08:00:15] <bahamat> https://wiki.smartos.org/pages/viewpage.action?pageId=755673
[08:00:24] <MagnusDredd_> It was super super cheap
[08:00:47] <arekinath> should work ok on a 1st-gen opteron, as long as you don't care for running KVM and don't have too much weird nforce hardware on the board
[08:00:52] <MagnusDredd_> that
[08:01:06] <MagnusDredd_> oops, that’s super useful info, thank you
[08:01:38] <MagnusDredd_> https://www.newegg.com/Product/Product.aspx?Item=N82E16813191001
[08:01:45] <arekinath> is this an X2100 or something? :P
[08:02:17] <arekinath> oh that board looks pretty ok
[08:02:32] <MagnusDredd_> Rackspace used tons of these
[08:02:50] <bahamat> MagnusDredd_: If you've got some money to spend, but don't want to break the bank, Intel NUCs are decent. I maintain a parts list with notes on which models are known to work http://a.co/9U9H7m6
[08:03:12] <arekinath> I dunno how the Sil3114 will go, but a bcm5702 should work, and the chipset should be ok in general
[08:03:13] <MagnusDredd_> I’ve got a pile of drives that I plan on throwing at it
[08:03:52] <MagnusDredd_> I’ve got some of these http://www.supermicro.com/products/accessories/addon/AOC-SAT2-MV8.cfm
[08:03:57] <MagnusDredd_> Also super cheap
[08:04:14] <MagnusDredd_> They’re basically what shipped in the Sun Thumper series machines
[08:04:25] <arekinath> oh, the marvell chips
[08:04:33] <bahamat> I use dual bay USB HDD enclosures. They're pretty cheap, and USB speeds are fast enough for my home needs.
[08:04:56] <arekinath> MagnusDredd_: we still have the old thumper drivers that might work with it if it's in that same marvell family
[08:05:19] <MagnusDredd_> I picked up two milk crates of hard drive enclosures and drives recently
[08:05:57] <MagnusDredd_> hmm, I think I will have to give SmartOS a try…
[08:06:20] <arekinath> looks like an 88SX6081, which marvell88sx should attach to
[08:06:25] <arekinath> it's on its PCI ID list anyway
[08:06:56] <MagnusDredd_> very nice
[08:07:02] <arekinath> if you want to see any of the disks on the Sil3114 onboard SATA on that board you'll probably have to chuck it into IDE emulation mode though
[08:07:04] <arekinath> just fyi
[08:07:20] <arekinath> I don't think we have a driver for it
[08:08:09] <MagnusDredd_> The version of the board I have does not have SATA onboard
[08:08:22] <arekinath> well then :)
[08:08:31] <MagnusDredd_> *chuckle*
[08:09:02] <MagnusDredd_> How little RAM will SmartOS deal with?
[08:09:25] <arekinath> we only ship a 64 bit kernel, and it'll barf on anything <1GB
[08:09:56] <MagnusDredd_> Cool, that’s completely reasonable
[08:09:59] <arekinath> plain smartos can run in 2GB VMs ok, but it really wants 4 or 8 minimum if you want to do more than a few small zones
[08:10:17] <arekinath> zfs ARC will always happily accept more DRAM
[08:10:24] <MagnusDredd_> I’ve got 2GB, which for serving files only has been ok
[08:10:25] <arekinath> it'll make use of whatever you can give it :P
[08:11:08] <arekinath> it's been a long time since I tried booting it on real hardware with <4GB, heh
[08:11:38] <MagnusDredd_> *nod*  I’m cheap…
[08:12:22] <MagnusDredd_> But the great thing about Solaris/FreeBSD is that you can throw it in the corner and mostly forget about it
[08:12:47] <arekinath> assuming you're not exposing it to the internet, yeah. :)
[08:13:29] <MagnusDredd_> absolutely, if I was there’s no way in hell I’d still be running OpenSolaris 2010.03 or otherwise
[08:14:12] <arekinath> haha sorry, just gotta mention it, you know.
[08:14:35] <MagnusDredd_> yes, definitely!
[08:15:14] <MagnusDredd_> The reason for asking about the Zones is that I’ve got a friend who does hosting on Linux and I help him out from time to time…
[08:15:18] <arekinath> smartos is designed to be easy to update the kernel and gz, since they're not installed on disk (just on the live usb/cd)
[08:15:34] <arekinath> you just reboot onto a new image and you're done
[08:16:08] <arekinath> but any software inside the zones like things from pkgsrc (or linux packages in lx) have to be managed separately
[08:16:48] <MagnusDredd_> …and he’s tied to one of those horrible web based management systems
[08:19:09] <arekinath> are any of those still under active development? webmin and friends?
[08:19:20] <MagnusDredd_> He’d like to use ZFS (send/recieve, snapshots, etc) but the overhead of extra VMs and the lack of granularity with files vs, having to set up NFS to another VM which supports ZFS properly...
[08:19:23] <arekinath> google tells me yes apparently
[08:19:24] <arekinath> huh
[08:19:34] <MagnusDredd_> There’s a bunch of them
[08:19:49] <arekinath> I feel like those are sealed away in a time capsule from 2003 in my head
[08:20:13] <MagnusDredd_> He uses Plesk which is horrible
[08:20:54] <MagnusDredd_> I’ve got to run
[08:21:11] <MagnusDredd_> you guys have been super helpful
[08:21:49] <MagnusDredd_> BTW, I must admit that I’m a big fan of Bryan Cantrill
[08:22:05] <MagnusDredd_> his BSD Now interviews are hysterical
[08:22:42] <arekinath> it's not hard to notice when he's in the office usually
[08:22:43] *** MagnusDredd_ <MagnusDredd_!~MagnusDre@wsip-98-172-66-146.ph.ph.cox.net> has left #smartos
[08:22:44] <arekinath> :P
[08:30:45] *** bahamat <bahamat!~bahamat@2001:470:f0f4:a0a:8000:ee1a:6aeb:f931> has quit IRC (Quit: Leaving.)
[08:32:06] *** bahamat <bahamat!~bahamat@2001:470:f0f4:a0a:8000:ee1a:6aeb:f931> has joined #smartos
[08:34:25] *** jhendricks1 <jhendricks1!~jhendrick@8.12.47.4> has quit IRC (Ping timeout: 240 seconds)
[08:59:59] *** s-mutin <s-mutin!~s-mutin@85.234.114.134> has joined #smartos
[09:01:53] *** mutin-s <mutin-s!~s-mutin@85.234.114.134> has quit IRC (Ping timeout: 248 seconds)
[09:33:49] *** Smithx10 <Smithx10!~smithx10@106.19.62.94.rev.vodafone.pt> has joined #smartos
[09:34:17] *** bahamat <bahamat!~bahamat@2001:470:f0f4:a0a:8000:ee1a:6aeb:f931> has quit IRC (Quit: Leaving.)
[09:34:38] *** Smithx10 <Smithx10!~smithx10@106.19.62.94.rev.vodafone.pt> has quit IRC (Remote host closed the connection)
[09:37:35] *** tozhu <tozhu!~Thunderbi@171.221.44.177> has quit IRC (Quit: tozhu)
[09:45:26] *** Smithx10 <Smithx10!~smithx10@106.19.62.94.rev.vodafone.pt> has joined #smartos
[09:46:12] *** Smithx10 <Smithx10!~smithx10@106.19.62.94.rev.vodafone.pt> has quit IRC (Remote host closed the connection)
[09:53:08] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Quit: Leaving.)
[10:01:01] *** raijin <raijin!~raijin@unaffiliated/raijin> has quit IRC (Ping timeout: 246 seconds)
[10:03:56] *** mnowak_ <mnowak_!~mnowak_@94.142.238.104> has joined #smartos
[10:14:26] *** Smithx10 <Smithx10!~smithx10@106.19.62.94.rev.vodafone.pt> has joined #smartos
[10:24:35] *** steven43226 <steven43226!~steven431@cpc106611-shef15-2-0-cust201.17-1.cable.virginm.net> has quit IRC (Ping timeout: 240 seconds)
[10:26:08] *** shyu <shyu!~shyu@114.250.79.206> has joined #smartos
[10:32:53] *** Smithx10 <Smithx10!~smithx10@106.19.62.94.rev.vodafone.pt> has quit IRC (Remote host closed the connection)
[10:48:06] *** raijin <raijin!~raijin@unaffiliated/raijin> has joined #smartos
[10:49:55] *** Smithx10 <Smithx10!~smithx10@106.19.62.94.rev.vodafone.pt> has joined #smartos
[10:52:35] *** Smithx10 <Smithx10!~smithx10@106.19.62.94.rev.vodafone.pt> has quit IRC (Remote host closed the connection)
[10:58:52] *** shyu <shyu!~shyu@114.250.79.206> has quit IRC (Ping timeout: 255 seconds)
[11:03:57] *** mnowak_ <mnowak_!~mnowak_@94.142.238.104> has quit IRC (Ping timeout: 240 seconds)
[11:14:44] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[11:17:51] *** KermitTheFragger <KermitTheFragger!~KermitThe@178-84-109-166.dynamic.upc.nl> has joined #smartos
[11:30:21] *** Smithx10 <Smithx10!~smithx10@106.19.62.94.rev.vodafone.pt> has joined #smartos
[11:31:38] *** Smithx10 <Smithx10!~smithx10@106.19.62.94.rev.vodafone.pt> has quit IRC (Remote host closed the connection)
[12:04:34] *** sjorge <sjorge!~sjorge@d5152645C.access.telenet.be> has quit IRC (Ping timeout: 255 seconds)
[12:05:01] *** sjorge <sjorge!~sjorge@d5152645C.access.telenet.be> has joined #smartos
[12:10:50] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Quit: Leaving.)
[12:11:24] *** andy_js <andy_js!~andy@94.3.225.134> has joined #smartos
[12:17:35] *** richlowe <richlowe!~richlowe@cpe-74-136-198-149.kya.res.rr.com> has quit IRC (Ping timeout: 240 seconds)
[12:18:41] *** Smithx10 <Smithx10!~smithx10@106.19.62.94.rev.vodafone.pt> has joined #smartos
[12:20:00] *** ed209 <ed209!~ed209@165.225.128.67> has quit IRC (Remote host closed the connection)
[12:20:06] *** ed209 <ed209!~ed209@165.225.128.67> has joined #smartos
[12:39:33] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[12:54:19] *** Shaun <Shaun!~quassel@203.161.108.48.static.amnet.net.au> has quit IRC (Ping timeout: 240 seconds)
[12:54:33] *** Shaun <Shaun!~quassel@203.161.108.48.static.amnet.net.au> has joined #smartos
[13:11:20] *** Smithx10 <Smithx10!~smithx10@106.19.62.94.rev.vodafone.pt> has quit IRC (Remote host closed the connection)
[13:14:34] *** KermitTheFragger <KermitTheFragger!~KermitThe@178-84-109-166.dynamic.upc.nl> has quit IRC (Remote host closed the connection)
[13:34:48] *** Tarogg <Tarogg!~HansDampf@ip4d168396.dynamic.kabel-deutschland.de> has joined #smartos
[13:39:47] *** Tarogg <Tarogg!~HansDampf@ip4d168396.dynamic.kabel-deutschland.de> has quit IRC (Quit: Leaving)
[13:42:08] *** richlowe <richlowe!~richlowe@2607:fcc8:bcd4:b100::2> has joined #smartos
[13:44:22] *** Smithx10 <Smithx10!~smithx10@106.19.62.94.rev.vodafone.pt> has joined #smartos
[13:44:31] *** Smithx10 <Smithx10!~smithx10@106.19.62.94.rev.vodafone.pt> has quit IRC (Remote host closed the connection)
[14:00:35] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Quit: Leaving.)
[14:00:44] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[14:05:05] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 240 seconds)
[14:10:21] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[14:16:33] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 248 seconds)
[14:20:43] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[14:24:57] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 240 seconds)
[14:30:04] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[14:31:09] *** Smithx10 <Smithx10!~smithx10@a109-49-149-179.cpe.netcabo.pt> has joined #smartos
[14:33:26] *** marsell <marsell!~Adium@114.23.152.33> has quit IRC (Quit: Leaving.)
[14:34:00] *** Smithx10 <Smithx10!~smithx10@a109-49-149-179.cpe.netcabo.pt> has quit IRC (Remote host closed the connection)
[14:34:44] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 260 seconds)
[14:36:37] *** marsell <marsell!~Adium@114.23.152.33> has joined #smartos
[14:38:54] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[14:43:36] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 260 seconds)
[14:45:08] *** Smithx10 <Smithx10!~smithx10@a109-49-149-179.cpe.netcabo.pt> has joined #smartos
[14:48:20] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[14:52:30] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 240 seconds)
[14:57:11] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[15:02:13] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 276 seconds)
[15:03:12] *** MagnusDredd_ <MagnusDredd_!~MagnusDre@wsip-98-172-66-146.ph.ph.cox.net> has joined #smartos
[15:06:37] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[15:09:11] *** marsell <marsell!~Adium@114.23.152.33> has quit IRC (Quit: Leaving.)
[15:09:24] *** irq_ <irq_!~irq@p5483C27D.dip0.t-ipconnect.de> has joined #smartos
[15:11:19] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 255 seconds)
[15:14:04] *** irq_ <irq_!~irq@p5483C27D.dip0.t-ipconnect.de> has quit IRC (Ping timeout: 255 seconds)
[15:15:28] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[15:16:45] *** jdnavarro <jdnavarro!~danny@200.111.121.146> has joined #smartos
[15:20:00] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 260 seconds)
[15:24:49] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[15:28:45] *** jdnavarro <jdnavarro!~danny@200.111.121.146> has quit IRC (Quit: WeeChat 1.7.1)
[15:29:46] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 268 seconds)
[15:29:47] *** Smithx10 <Smithx10!~smithx10@a109-49-149-179.cpe.netcabo.pt> has quit IRC (Read error: Connection reset by peer)
[15:30:10] *** Smithx10 <Smithx10!~smithx10@a109-49-149-179.cpe.netcabo.pt> has joined #smartos
[15:33:41] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[15:39:07] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 246 seconds)
[15:42:55] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[15:47:05] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 240 seconds)
[15:52:16] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[15:56:27] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 240 seconds)
[15:57:28] <bennabiy> what is the best way to set up a private network from VM to VM within the same GZ and is it possible for the connection to simulate a 10Gig network connection?
[15:58:39] <bennabiy> So basically I want the traffic to stay within the server, and not leave on the line just to come back to itself
[15:58:47] <bennabiy> but from one zone to another
[16:01:08] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[16:05:48] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 258 seconds)
[16:10:20] *** Smithx10 <Smithx10!~smithx10@a109-49-149-179.cpe.netcabo.pt> has quit IRC (Remote host closed the connection)
[16:10:33] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[16:10:57] *** Smithx10 <Smithx10!~smithx10@a109-49-149-179.cpe.netcabo.pt> has joined #smartos
[16:15:24] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 268 seconds)
[16:19:24] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[16:24:02] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 268 seconds)
[16:28:16] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[16:30:31] *** Smithx10_ <Smithx10_!~smithx10@a109-49-149-179.cpe.netcabo.pt> has joined #smartos
[16:30:32] *** Smithx10 <Smithx10!~smithx10@a109-49-149-179.cpe.netcabo.pt> has quit IRC (Read error: Connection reset by peer)
[16:32:46] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 255 seconds)
[16:37:54] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[16:41:57] *** tozhu <tozhu!~Thunderbi@101.207.125.86> has joined #smartos
[16:42:15] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 255 seconds)
[16:44:16] <rmustacc> bennabiy: Use an etherstub (nictagadm -l)
[16:44:48] <bennabiy> rmustacc: and then how do I associate that with each VM?
[16:46:26] <rmustacc> Via the nictag.
[16:46:28] <bennabiy> I am sorry, my brain is somewhat fried right now after a long night, I am sure that info is on the wiki
[16:46:43] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[16:47:11] <bennabiy> ok, and what speeds would I get over that link?
[16:47:14] <rmustacc> You use nictagadm -l (see the manual page) to create a nic tag based on an etherstub.
[16:47:31] <rmustacc> There's no fixed speed, because its virtual
[16:47:36] <bennabiy> wonderful
[16:47:53] <bennabiy> what is the status of vmadm send / receive?
[16:51:07] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 260 seconds)
[16:52:36] *** tozhu <tozhu!~Thunderbi@101.207.125.86> has quit IRC (Quit: tozhu)
[16:53:48] *** Gathis <Gathis!~TheBlack@unaffiliated/gathis> has joined #smartos
[16:54:05] <bennabiy> rmustacc: would there be a benefit of bumping the MTU to 9000 for the etherstub?
[16:54:20] <bennabiy> for sending large files
[16:55:35] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[16:56:44] <bennabiy> I guess I cannot specify it :)
[16:59:45] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 248 seconds)
[17:02:24] <bennabiy> rmustacc: one more question, and I am sure it is simple. What is the best way to add that new interface to the vm? would I need to pass the JSON through vmadm?
[17:02:38] <bennabiy> and is it best to do it with the vm stopped?
[17:05:15] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[17:09:27] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 240 seconds)
[17:14:05] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[17:18:32] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 260 seconds)
[17:22:57] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[17:27:16] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 255 seconds)
[17:27:51] <bennabiy> I guess Sunday is the wrong day to have problems
[17:29:17] * bennabiy begs mercy of anyone who might be around.
[17:30:06] *** Smithx10_ <Smithx10_!~smithx10@a109-49-149-179.cpe.netcabo.pt> has quit IRC (Remote host closed the connection)
[17:31:03] <wiedi> bennabiy: check out "example 9" in the vmadm man page
[17:31:07] <bennabiy> when I add an etherstub to a VM, do I need to add a MAC to it?
[17:31:38] <bennabiy> wiedi: I more have problems with migrating zones
[17:32:21] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[17:32:36] <bennabiy> I am moving a zone from one GZ to another, and I need to move it from a KVM to an LX zone, and need to still have encryption at rest in that zone
[17:33:20] <bennabiy> I currently have my kvm zvols cloned / cloning over to the other GZ
[17:33:49] <bennabiy> I have the .xml file for the kvm in place. I added an etherstub called vmswitch0 to the GZ
[17:34:47] <wiedi> since openzfs encryption is not merged yet you might be able to do something with lofiadm, but I don't know how that works and have never done it
[17:34:47] <bennabiy> I am about to do something along the lines of add_nics: to the KVM and the LX zone to put them on a common VM only network for sending the dataset from kvm zvol to lx dataset
[17:35:41] <bennabiy> yes. I was looking at lofiadm, but had a question of whether I would be able to expand the space if I needed once I encrypted the data?
[17:36:40] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 255 seconds)
[17:36:55] <bennabiy> Like, could I create a dataset for use in the lx zone, and then if it gets full, add to it somehow another backing file?
[17:38:46] <wiedi> I guess you'd create a zpool on top of the lofi device, and then you can add more lofi devices to your pool
[17:39:10] <bennabiy> and how would I delegate the space in that pool to the vm?
[17:39:18] <bennabiy> I found http://constantin.glez.de/blog/2012/02/introducing-sparse-encrypted-zfs-pools
[17:39:25] <bennabiy> which is somewhat what I want to do
[17:41:13] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[17:45:35] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 240 seconds)
[17:45:36] <wiedi> you can specify the zpool property for vmadm create
[17:46:29] *** bahamat <bahamat!~bahamat@2001:470:f0f4:a0a:8000:ee1a:6aeb:f931> has joined #smartos
[17:50:41] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[17:54:55] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 240 seconds)
[17:59:32] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[18:03:57] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 240 seconds)
[18:04:40] *** Smithx10 <Smithx10!~smithx10@106.19.62.94.rev.vodafone.pt> has joined #smartos
[18:08:58] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[18:13:22] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 260 seconds)
[18:17:50] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[18:22:25] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 248 seconds)
[18:26:41] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[18:28:11] <flatline> benably: do you not have an admin nictag and an extenal nictag?
[18:29:05] <flatline> if admin is not routed you could send using it...
[18:31:20] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 260 seconds)
[18:36:02] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[18:39:46] <bennabiy> right now admin is external
[18:40:14] <bennabiy> wiedi: so to understand you, I can specify which zpool to use for dataset?
[18:40:37] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 255 seconds)
[18:44:53] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[18:49:34] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 255 seconds)
[18:52:54] <bennabiy> can I use delegated_dataset on an lx zone?
[18:54:14] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[18:54:35] <arekinath> bennabiy: yes, but note that to manage it you'll need to use the zfs tools in /native, not the tools from zfsonlinux
[18:54:54] <arekinath> we don't emulate the zfsonlinux control device at the moment
[18:55:40] <arekinath> (they changed a few ioctls and structs)
[18:56:32] <bennabiy> ok
[18:58:04] <bennabiy> I just read the man for vmadm and did not see the proper syntax for adding a dataset from another pool to the lx zone
[18:58:27] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 240 seconds)
[18:58:46] <arekinath> it'll delegate from the pool the zone's normal dataset is on
[18:59:14] <arekinath> if you need to delegate something that's not on the same pool as the zone root you can do it, but not through vmadm
[18:59:41] <arekinath> you'll have to go behind it's back and use zonecfg
[18:59:50] <arekinath> *its
[19:00:46] <bennabiy> I am going to be making a lofs backed encrypted zpool and would like to use that for a folder within the zone.
[19:02:02] <arekinath> well, you're pretty far off vmadm's normal rails there no matter how you want to do it at the moment :)
[19:02:25] <arekinath> we've been waiting for openzfs encryption to land before supporting anything
[19:02:48] <bennabiy> yes, and I look forward to that
[19:03:06] <bennabiy> but I would like to get away from the layer upon layer upon layer to get encryption
[19:03:07] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[19:03:41] <LeftWing> For interest's sake, why is encryption a requirement here?
[19:04:22] <bennabiy> Secure email server / groupware
[19:04:26] <bennabiy> primarily
[19:04:40] <bennabiy> encryption for at rest
[19:07:10] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 240 seconds)
[19:08:01] <arekinath> you could maybe delegate full control of lofi and zfs to a zone and do it entirely inside it
[19:08:17] <bennabiy> lx zone as well?
[19:08:35] <bennabiy> right now this needs to be lx, I would like to get to native, but am not there yet
[19:08:40] <arekinath> yes, just once again you'll have to use tools in /native
[19:08:50] <bennabiy> no problems with that
[19:09:02] <bennabiy>  /native does not scare me
[19:09:09] <arekinath> That zone will also have enough access to wreck the whole box
[19:09:26] <arekinath> if it's delegated enough privs to mount its own zfs
[19:10:01] <bennabiy> No partial delegation?
[19:10:02] <arekinath> well, its own zpool, really
[19:10:33] <arekinath> you can't delegate it the right to manage entire pools without also giving it the right to manage the main "zones" pool
[19:10:34] <bennabiy> if I had a zone which was based on an lx image, but then had a zpool allocated to it
[19:10:46] *** bahamat <bahamat!~bahamat@2001:470:f0f4:a0a:8000:ee1a:6aeb:f931> has quit IRC (Quit: Leaving.)
[19:10:50] <bennabiy> hrm
[19:10:54] <arekinath> it's all or nothing when it comes to *pool* management
[19:11:05] <arekinath> delegated datasets are only a single dataset
[19:11:15] <arekinath> you don't get any rights over the pool with them
[19:11:16] <bennabiy> so keep the lofi at the GZ level and give it that as a dataset?
[19:11:28] <arekinath> that's an option, too, yeah
[19:11:53] <arekinath> do the lofiadm and zpool import in a SMF service transient script in the gz before zones start
[19:12:29] <arekinath> and use zonecfg to delegate a piece of the inner pool (or even its root dataset) to the zone
[19:12:40] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[19:12:59] <bennabiy> any example of the proper syntax for that?
[19:13:26] <bennabiy> tgerczei was kind enough to give me an example script, but I still need to know how to initially set up the zone
[19:13:46] <bennabiy>  I am creating the zone now, and have my kvm moved over so I can sync the data out of it, and into the lx zone
[19:14:26] <arekinath> http://docs.oracle.com/cd/E19253-01/819-5461/gbbst/index.html is still correct for us
[19:14:32] <bennabiy> The lx zone creation, and the lofiadm part have not been done yet, so I am all ears
[19:14:39] <bennabiy> arekinath: are you with joyent?
[19:14:43] <arekinath> you'll have to do that after creating the zone with vmadm
[19:15:23] <arekinath> bennabiy: yeah
[19:15:35] <tgerczei> my script, however, isn't SMF-aware nor quite ready, although I could make it a service with minimal effort
[19:16:19] <tgerczei> let me know if you need a hand there
[19:16:37] <arekinath> you can just make a transient SMF service that executes an arbitrary script during boot, more or less
[19:16:40] <bennabiy> arekinath: thank you. From what you linked, it looks like the NGZ only has access to the dataset from what it is told on down, not to manage anything above it
[19:16:42] <arekinath> just needs a bit of XML
[19:16:50] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 240 seconds)
[19:16:58] <arekinath> And then you chuck it in /opt/custom/smd
[19:17:02] <arekinath> *smf
[19:17:11] <bennabiy> tgerczei: I need all the help I can get. I am on a time crunch and my main mail server is down due to full drive on the current server
[19:17:24] <arekinath> bennabiy: correct. That's what a delegated dataset is
[19:18:10] <bennabiy> arekinath: I do not mind my zone being able to manage the dataset I give it. As long as it cannot wreck other datasets outside of itself
[19:19:00] <arekinath> right, a delegated dataset gives it no ability to directly touch datasets outside the children of the one you gave it
[19:19:19] <arekinath> keep in mind zfs recv is still disabled by default for ngzs to preserve that property
[19:19:30] <bennabiy> <arekinath> That zone will also have enough access to wreck the whole box
[19:19:30] <bennabiy> <arekinath> if it's delegated enough privs to mount its own zfs
[19:19:36] <bennabiy> that was why I wanted to make sure
[19:19:51] <arekinath> I meant pool there not zfs
[19:19:54] <arekinath> sorry
[19:20:01] <arekinath> corrected it further down
[19:20:01] <bennabiy> ok
[19:20:18] <bennabiy> zfs send works though, correct?
[19:20:29] <arekinath> yes, send is fine
[19:20:45] <bennabiy> can a zfs recv be done from the GZ?
[19:20:47] <arekinath> recv can be used to write arbitrary data into the pool and cause problems though
[19:20:51] <arekinath> so it's disabled
[19:20:59] <arekinath> the gz can do recv, yea
[19:21:02] <bennabiy> ok
[19:21:31] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[19:24:11] <tgerczei> bennabiy: query me with what you need help with, if you need me
[19:24:51] *** bahamat <bahamat!~bahamat@ip72-199-207-63.sd.sd.cox.net> has joined #smartos
[19:25:35] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 240 seconds)
[19:26:06] <bennabiy> tgerczei: Is it better to set up the lofiadm portion before I even create the lx zone, or can I add that in after?
[19:29:35] <arekinath> you can add it in later right now while you're putting it together
[19:29:44] <arekinath> but at boot it'll need to be done first
[19:29:57] <tgerczei> yeah, sorry for coming late to the party
[19:30:28] <bennabiy> I wasn't sure if it would be better to have the zpool ready while creating the vm, or if it is fine to add it in after vm creation
[19:30:46] <bennabiy> I am editing the lx zone json getting ready to import it now
[19:30:51] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[19:31:00] <bennabiy> err creat
[19:31:01] <bennabiy> e
[19:31:01] <tgerczei> I did it afterwards but I use zonecfg instead of vmadm and delegation in there
[19:31:04] <arekinath> you can add it in later, if the zone is already running it'll just need to be restarted to pick up its delegated dataset
[19:31:32] <tgerczei> and I don't use an LX-brand zone for my use-case, rather a simple native one
[19:31:46] <tgerczei> that shouldn't really matter at this stage though
[19:34:41] <bennabiy> ok, so I will just create the zone, and then add the dataset after
[19:35:03] <bennabiy> I found that delegated datasets only worked if added at creation with vmadm, not so with zonecfg?
[19:35:16] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 260 seconds)
[19:37:27] <arekinath> bennabiy: should work fine with zonecfg, as long as the dataset path you give it already exists, and you reboot the zone after the change
[19:37:50] <arekinath> bennabiy: don't forget to use the "commit" command in zonecfg after making changes, too
[19:38:08] <bennabiy> arekinath: ok. Oh? I have never used zonecfg so thank you!
[19:38:18] <arekinath> vmadm internally is just running zonecfg fwiw
[19:38:24] <bennabiy> I noticed that
[19:38:41] <bennabiy> In some debugging we were doing yesterday
[19:39:04] *** jrick <jrick!~jrick@unaffiliated/jrick> has quit IRC (Ping timeout: 255 seconds)
[19:39:42] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[19:40:26] <bennabiy> so, say I did add dataset in zonecfg and set name=2ndzpool/zone/foobar how would I reference that from within the zone? as 2ndzpool/zone/foobar?
[19:40:54] *** jrick <jrick!~jrick@unaffiliated/jrick> has joined #smartos
[19:41:39] <bennabiy> with lofiadm I am making the devices, and with zpool I am making the second pool in the GZ, and then zfs adding a dataset to the second zpool, and then zonecfg adding that zfs dataset to the zone?
[19:41:43] <bennabiy> do I get it?
[19:42:13] <bennabiy> and within the zone, I use /native/zfs to set mountpoint for that dataset
[19:42:14] <bennabiy> etc
[19:42:26] <bennabiy> arekinath, tgerczei : ^
[19:42:31] <arekinath> yep
[19:42:54] <bennabiy> and I would refer to it as the same name I used in the setting up the delegation?
[19:43:22] <arekinath> yes, it keeps the same full name inside the zone
[19:44:10] <bennabiy> so if my lofs pool is tanker, and my dataset on it is oil, then within the zone I would reference tanker/oil and snapshots within the zone would be tanker/oil@blah right?
[19:44:22] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 260 seconds)
[19:44:33] <bennabiy> I am just really wanting to make sure I understand it
[19:44:40] <arekinath> yes
[19:44:43] <bennabiy> ok
[19:44:56] <arekinath> and by default it'll mount at /tanker/oil in the zone
[19:45:03] <arekinath> but you can change that
[19:45:13] <bennabiy> from within the zone correct?
[19:45:26] <arekinath> sometimes you have to play with LX startup scripts to make mounting work on them, depending on distro
[19:45:29] <arekinath> yeah
[19:45:46] <bennabiy> this is not something which is required for booting, so that helps
[19:46:52] <bennabiy> now, if I make say, 10 files at 10GB each, and find I run out of room, or am getting tight, and I make 10 more, how do I add them to the zpool?
[19:47:19] <bennabiy> and what will I need to do to give the delegated dataset access to that boost?
[19:47:30] <arekinath> you'll have to make lofi devices for them in the gz and do the zpool add in the gz
[19:48:01] <bennabiy> arekinath: now, if these are backed on a mirrored zpool, do I need to do a mirrored lofi zpool?
[19:48:10] <arekinath> no
[19:48:13] <bennabiy> ok
[19:48:24] <arekinath> I would keep the inner pool to just single device vdevs
[19:48:25] <bennabiy> so the base mirror is covering for datacorruption
[19:48:39] *** bahamat <bahamat!~bahamat@ip72-199-207-63.sd.sd.cox.net> has quit IRC (Quit: Leaving.)
[19:48:45] <bennabiy> just making sure
[19:48:45] <arekinath> yeah
[19:49:30] <arekinath> also I would put lz4 compression on on the inner pool
[19:49:40] <arekinath> especially if its dbs of email and stuff
[19:49:49] <arekinath> it'll compress well for cheap
[19:49:56] <bennabiy> I have lz4 on the whole zpool
[19:50:14] <arekinath> right, but to the outer pool the encrypted data is incompressible
[19:50:16] *** tozhu <tozhu!~Thunderbi@101.207.125.86> has joined #smartos
[19:50:20] <arekinath> it looks like random noise
[19:50:20] <bennabiy> oh, that is right
[19:50:26] <bennabiy> good call
[19:50:40] <arekinath> I would turn lz4 off on the outer pool dataset where the lofi files are
[19:50:45] <bennabiy> that is why you get paid to do this, and I don't :)
[19:50:49] <arekinath> because it won't be able to compress them
[19:51:05] <bennabiy> ok,
[19:51:29] *** tozhu <tozhu!~Thunderbi@101.207.125.86> has quit IRC (Client Quit)
[19:51:41] <bennabiy> would you recommend setting up a zfs dataset off zones pool to host the lofs files
[19:51:50] *** ismell <ismell!~ismell@2603:300b:c0d:ea00:20e:4ff:fe24:4477> has quit IRC (Ping timeout: 240 seconds)
[19:52:08] <arekinath> yeah, I would put them in their own dataset so you can track their space usage and such easily
[19:52:15] <bennabiy> wise
[19:52:16] <bennabiy> ok
[19:52:21] <arekinath> and snapshot them potentially if you need to sometime
[19:52:35] <arekinath> you might not need it but it can be useful
[19:52:40] <bennabiy> yes
[19:52:48] <bennabiy> I always like fallbacks for those things
[19:52:48] <arekinath> and you can give them different policy for compression that way too
[19:53:50] <bennabiy> heh, I already made one (I just looked at my list and there it was ) :)
[19:54:01] <bennabiy> no files in it though
[19:54:16] <bennabiy> arekinath: would you recommend lots of smaller files, or chunks of bigger files?
[19:54:27] <bennabiy> I guess it would only matter when it comes to migrating it
[19:56:01] <arekinath> probably bigger files, I can't remember how many lofi devices you can have at once but I suspect it's not a huge number
[19:56:25] <bennabiy> Also, what is the best way to generate the files?
[19:56:26] <bennabiy> dd?
[19:57:06] <bennabiy> or mkfile?
[19:57:38] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[19:58:31] <arekinath> I don't think it matters. lofiadm man page recomments mkfile
[20:02:09] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 248 seconds)
[20:03:36] <bennabiy> hrm, lofiadm man does not specify upper limit of how many devices it can have
[20:06:59] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[20:09:26] <arekinath> bennabiy: yeah, but there's only so many bits in the minor number it'll use, because of masking... I suspect you can only have maybe 1024 (10 bits) or something of them... would have to read a lot of code to figure it out since it all depends on what CMLB is doing
[20:09:51] <bennabiy> as long as it is not like 10 :)
[20:10:27] <arekinath> it's not like 10 :P
[20:10:58] <bennabiy> will vmadm automatically assign my vm a mac address for use with an etherstub?
[20:11:12] <bennabiy> if I use the etherstub nic_tag
[20:11:25] <arekinath> yes, vnics in general get a generated MAC by default
[20:11:32] <arekinath> on etherstubs or real intfs
[20:11:35] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 268 seconds)
[20:11:59] <bennabiy> ok
[20:12:15] <bennabiy> so I should only need an entry for nic_tag and ips?
[20:12:51] <arekinath> yes
[20:16:19] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[20:19:43] *** mnowak_ <mnowak_!~mnowak_@94.142.238.104> has joined #smartos
[20:20:31] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 246 seconds)
[20:25:09] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[20:25:16] <bennabiy> and the primary should be the one I need to be public facing?
[20:29:16] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 246 seconds)
[20:29:19] <bennabiy> I got an invalid IP for NIC
[20:29:24] <bennabiy> on my etherstub
[20:29:37] <bennabiy> 10.0.10.10
[20:31:24] <bennabiy> why would it tell me that?
[20:34:34] *** tpaul <tpaul!~travispau@59.12.59.18> has joined #smartos
[20:35:48] <bennabiy> Can someone help me with this etherstub issue?
[20:36:15] <bennabiy> do I need to do something once I create my etherstub to allow a VM to use it?
[20:37:22] <bennabiy> https://pastebin.mozilla.org/9027836 is my current json for nics
[20:37:47] <arekinath> "ips": needs to be an array
[20:37:52] <arekinath> make it ["10.0.10.10"]
[20:38:01] <bennabiy> I tried that at first
[20:38:04] <arekinath> or just "ip":
[20:38:05] <bennabiy> same thing
[20:38:08] <arekinath> since you're giving a netmask
[20:38:16] <arekinath> "ips": has to take cidr
[20:38:22] <arekinath> so it would have to be ["10.0.10.10/24"]
[20:38:22] <bennabiy> ah
[20:38:30] <arekinath> and no netmask
[20:38:30] <bennabiy> makes sense
[20:38:57] *** tpaul <tpaul!~travispau@59.12.59.18> has quit IRC (Ping timeout: 248 seconds)
[20:39:18] <bennabiy> great
[20:39:23] <bennabiy> that was what I was missing
[20:39:36] *** irq_ <irq_!~irq@p5483C27D.dip0.t-ipconnect.de> has joined #smartos
[20:39:50] <bennabiy> and to add that etherstub to an existing KVM, I would need to use add_nics?
[20:40:04] <arekinath> yes
top

   July 23, 2017
< | 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