https://pulumi.com logo
Join Slack
Powered by
pulumi:pulumi:Stack: operations-Babylon-4 info...
# general
s
pulumipulumiStack: operations-Babylon-4 info: panic: runtime error: invalid memory address or nil pointer dereference [signal 0xc0000005 code=0x0 addr=0x0 pc=0x128929d]
😙 2
s
Hi @sparse-insurance-40223! That doesn’t look so good... do you have the backtrace that came from the panic available?
s
sorry no 😞
all the trace was within the kubernetes bits
s
Ah - thanks for reporting it. That’s likely one for @creamy-potato-29402 when he’s online.
c
Not good. Will fix ASAP.
@sparse-insurance-40223 I’m not quite sure what you mean when you say that it didn’t delete the primary resource group? You mean it tried and failed?
s
Yeah I think it left something behind and then got itself in a tizzy and the local state was corrupted or something
c
Did it leave an error?
Also what version of pulumi-kubernetes are you running?
This stack trace makes no sense lol
Also can you reproduce it
s
lol :S
how do i get the version
c
Just look at the NPM version of the dependency
s
resource-kubernetes-v0.17.0-rc1
oh i see one sec
c
hmm
was this coming from an example?
s
"@echoing-match-29901/kubernetes": "v0.17.0-rc1",
oh great thanks slack
c
lol
s
i was following an example yeah but had changed it pretty heavily by that point
basically scripted rg + k8s + nginx ingress + some permissions
and when i did pulumi destroy i got that panic
the only thing left in azure was the original RG (not the k8s auto-rg it makes for the cluster)
c
hmm, so did you run 
npm install
before sunday?
s
yeah i started on it on friday
c
makes sense. We weren’t publishing correct dev versions until sunday.
(We’re on v0.17.1-dev)
s
ah
well i'm planning on destroying this cluster tomorrow i can see if it happens again then
c
Hmm, so it looks like it’s calling 
Read
during refresh, then failing because there is no live object.
s
ah
c
Trying to get a repro… we should be catching the live object if it doesn’t exist.
s
i had a few more issues presumably relating to the confused local state, i think i ended up deleting the state folder
but they were all things like host doesnt exist cos the cluster had gone
c
oh
you were using a local statefile?
s
yeah
c
mmm got it
s
ah i didnt think to mention that 🙂
c
Hmm, does it reproduce reliably?
s
not sure tbh
i'm going to be deleting this test cluster tomorrow probably i can try then and get more info for the bug report
c
Ok well
please do try to reproduce it… I can’t get it to fail in quite this way.
I’ll add more detail.
s
hometime here, 6pm! i'll let you know how i get on tomorrow
ok will do
c
@sparse-insurance-40223 I think I understand the bug: https://github.com/pulumi/pulumi/issues/1984#issuecomment-424428743
if we can get a repro the fix should be trivially easy, I just want to make sure it is what we think. 🙂
@sparse-insurance-40223 I’ve developed a repro, and fixed the issue in https://github.com/pulumi/pulumi-kubernetes/pull/221
🙂
s
nice!
Open in Slack
PreviousNext
Powered by