rsalveti has quit [Quit: Connection closed for inactivity]
bgilbert has quit [Ping timeout: 256 seconds]
jpn has joined #fedora-coreos
jpn has quit [Ping timeout: 256 seconds]
paragan has quit [Quit: Leaving]
gursewak has quit [Ping timeout: 244 seconds]
gursewak has joined #fedora-coreos
jcajka has joined #fedora-coreos
azukku has joined #fedora-coreos
gursewak has quit [Ping timeout: 244 seconds]
paragan has joined #fedora-coreos
jpn has joined #fedora-coreos
jpn has quit [Ping timeout: 276 seconds]
jcajka has quit [Quit: Leaving]
gursewak has joined #fedora-coreos
cyberpear has quit [Quit: Connection closed for inactivity]
jcajka has joined #fedora-coreos
Betal has quit [Quit: WeeChat 3.5]
jpn has joined #fedora-coreos
jpn has quit [Ping timeout: 240 seconds]
wolfshappen has quit [Quit: later]
nemric has joined #fedora-coreos
nemric has quit [Quit: WeeChat 3.5]
nemric has joined #fedora-coreos
wolfshappen has joined #fedora-coreos
jpn has joined #fedora-coreos
wolfshappen has quit [Quit: later]
wolfshappen has joined #fedora-coreos
dwalsh has joined #fedora-coreos
heldwin has joined #fedora-coreos
jpn has quit [Quit: Lost terminal]
jpn has joined #fedora-coreos
jpn has quit [Quit: Lost terminal]
jpn has joined #fedora-coreos
jpn has quit [Client Quit]
jpn has joined #fedora-coreos
io7m has joined #fedora-coreos
io7m_ has joined #fedora-coreos
io7m has quit [Quit: Client closed]
fifofonix has joined #fedora-coreos
jcajka has quit [Quit: Leaving]
paragan has quit [Ping timeout: 256 seconds]
dwalsh has quit [Ping timeout: 248 seconds]
jcajka has joined #fedora-coreos
dwalsh has joined #fedora-coreos
jakfrost has joined #fedora-coreos
jakfrost has quit [Quit: Leaving]
ravanelli has joined #fedora-coreos
crobinso has joined #fedora-coreos
saqali has joined #fedora-coreos
mheon has joined #fedora-coreos
sayan has quit [Ping timeout: 244 seconds]
sayan has joined #fedora-coreos
<dustymabe>
jlebon: here's the NO_UPLOAD PR I submitted yesterday: https://github.com/coreos/fedora-coreos-pipeline/pull/547 - that one should allow us to debug the UEFI testiso issue in the pipeline without uploading a bunch of results IIUC
azukku has quit [Remote host closed the connection]
plarsen has joined #fedora-coreos
mnguyen has joined #fedora-coreos
jcajka has quit [Quit: Leaving]
mnguyen_ has quit [Ping timeout: 240 seconds]
dwalsh_ has joined #fedora-coreos
dwalsh has quit [Ping timeout: 246 seconds]
dwalsh__ has joined #fedora-coreos
dwalsh_ has quit [Ping timeout: 256 seconds]
dwalsh_ has joined #fedora-coreos
dwalsh__ has quit [Ping timeout: 256 seconds]
nemric has quit [Quit: WeeChat 3.5]
<ravanelli>
I was thinking about starting the fcos release today, if that's ok for you dustymabe?
<dustymabe>
ravanelli: would prefer to wait til monday if possible. We sometimes start it on Friday for good reasons, but we need to announce it and such. We've started it on Friday before and had some people not happy about it. We try to stick to this contract: https://github.com/coreos/fedora-coreos-tracker/issues/571
<ravanelli>
dustymabe: aaa ok, that's fine ;)
<lucab>
dustymabe: wrt generators and systemd, I think a better approach could be to make the idempotent (so they either do nothing or refresh the generated content when re-run)
<lucab>
*make them
io7m_ has left #fedora-coreos [#fedora-coreos]
Betal has joined #fedora-coreos
io7m has joined #fedora-coreos
<io7m>
testing...
<io7m>
*ahem*, excuse me. sorry, i wasn't registered and thus not allowed to speak :)
dwalsh__ has joined #fedora-coreos
dwalsh_ has quit [Ping timeout: 256 seconds]
<dustymabe>
lucab: yeah, but that's quite a bit more work to convert them all. in reality especially in the initrd, nothing should have changed
<io7m>
the ignition file does succeed in that i do get a user created with the right password and SSH key installed, but there's just no networking configured
<io7m>
i'm unsure if i should even be expecting there to be any
<dustymabe>
so you can SSH in?
<dustymabe>
or are you using the VGA console?
<io7m>
the VGA console. just to check that there's a user there and that keys exist
<dustymabe>
ok, so you logged in with the username/password on the VGA console
<io7m>
yep
<dustymabe>
if you run `ip link show` do you see interfaces listed?
<io7m>
there are interfaces listed, but none of them have addresses assigned
<dustymabe>
what does `nmcli c show` tell you?
<dustymabe>
any connections up?
<io7m>
one sec, let me get the machines up again... (waits for terraform to apply)
bgilbert has joined #fedora-coreos
jpn has quit [Ping timeout: 240 seconds]
<io7m>
dustymabe: i'm sure you can guess what i'm going to say. i just kicked off some new instances... they've each come up with working networking
<dustymabe>
\o/
<dustymabe>
maybe some DHCP race condition?
<io7m>
it's possible! my understanding is that networking at digitalocean is kind of ... strange. they don't support ipv6 for custom images, for example
<io7m>
vultr have no problems doing that
<dustymabe>
yeah - i guess keep watching to see if the behavior continues or not
<dustymabe>
hoping you don't see it again
<dustymabe>
do you mind updating the discussion forum post?
<io7m>
will do!
<io7m>
this wasn't a one-off occurrence, earlier. i destroyed and recreated the instances several times whilst playing with terraform
dwalsh_ has joined #fedora-coreos
dwalsh__ has quit [Ping timeout: 246 seconds]
jpn has joined #fedora-coreos
jpn has quit [Ping timeout: 244 seconds]
arnulfo_7 has joined #fedora-coreos
dwalsh__ has joined #fedora-coreos
dwalsh_ has quit [Ping timeout: 248 seconds]
jpn has joined #fedora-coreos
ravanelli has quit [Remote host closed the connection]
arnulfo_7 has quit [Ping timeout: 276 seconds]
<dustymabe>
gursewak: want to get together to debug the live iso failure in the pipeline?
arnulfo_7 has joined #fedora-coreos
jpn has quit [Ping timeout: 260 seconds]
<dustymabe>
random aside: rpm-ostree kargs feels way more heavyweight than it should be
<dustymabe>
bgilbert jlebon - i wonder what you think of my comment about terraform from 30m ago
<bgilbert>
dustymabe: this is a proposal that comes up periodically
<dustymabe>
obviously not going to change anything today, but we could decide on something like that for future changes
<dustymabe>
bgilbert: ahh, didn't know it had been previously considered
<bgilbert>
dustymabe: the problem with terraform is that it's a bunch of code we don't control, and we want to do very specific cloud things with fine-grained control
<bgilbert>
dustymabe: maybe it's better these days, but it also didn't have a great reputation for reliability back in ze day
<dustymabe>
bgilbert: i wonder if there is a balance we could maintain
<dustymabe>
i.e. use terraform for stuff where it works and use SDKs when we need finer grained control
<bgilbert>
dustymabe: maybe? but honestly that mantle code doesn't change much these days, so it doesn't really seem worth the upheaval
<dustymabe>
i've honestly never really dug into terraform too too much, but I get the idea
<dustymabe>
bgilbert: right, but if we get to the point where we're adding new platforms and such.. i think it would be worth considering the alternative
<dustymabe>
right now we're trying to update azure libraries and it's a bit of a mess
<bgilbert>
eh. I won't hard-block it, but I think the current approach has served us pretty well
<bgilbert>
someone would certainly need to do an experimental port before we committed to anything
<dustymabe>
the other approach is using the CLIs
<dustymabe>
i.e. instead of calling ore we call gcloud or azure cli or aws cli
<dustymabe>
but the integration with the SDKs is nice for kola
mnguyen_ has joined #fedora-coreos
mnguyen has quit [Ping timeout: 246 seconds]
<jlebon>
bgilbert: stamped
<jlebon>
mostly agreed with bgilbert re. terraform
<jlebon>
the CLI idea is interesting though. it might be worth it even if we have to use other parts of the sdk anyway for kola.
<bgilbert>
I guess the thing that's changed there is that we now assume that kola runs in the cosa container
<bgilbert>
the original design could not assume anything about the host system
ravanelli has quit [Remote host closed the connection]
jpn has quit [Ping timeout: 276 seconds]
dwalsh__ has quit [Ping timeout: 260 seconds]
<io7m>
so apparently today's the day i encounter edge cases for various cloud providers. got a different odd issue now with vultr. i can use the vultr-cli command-line tool and check that an instance has the expected ignition JSON embedded in the user_data field, but the machine says there's no ignition config provided on boot
<io7m>
additionally, the fedora coreos web site delivers an image in a format that vultr don't support (raw.xz), so i have to use the bare metal iso
<bgilbert>
io7m: is there a mechanism for creating images directly from a remote URL?
<io7m>
bgilbert: there is, yeah, but it won't accept xz compressed data. i'm in the process of decompressing and uploading the vultr .xz image to see if it behaves differently
<bgilbert>
I expect your problem will be fixed by using the vultr image, but for the record:
<bgilbert>
you're seeing two issues. 1) Ignition thinks it's booting on bare metal, so it doesn't know to read vultr userdata, and 2) the Ignition config passed to the live ISO only controls the live ISO boot, and is separate from the Ignition config passed to the installed system.
<io7m>
right!
<bgilbert>
(ignore (2) if you planned to run FCOS directly from the ISO, rather than installing)
<io7m>
purely experimenting at the moment (hence trying both digitalocean and vultr)
mheon has quit [Ping timeout: 276 seconds]
<bgilbert>
+1
dwalsh__ has joined #fedora-coreos
ccha has quit [Quit: WeeChat 3.5]
fifofonix has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
ccha has joined #fedora-coreos
crobinso has quit [Remote host closed the connection]