-
Notifications
You must be signed in to change notification settings - Fork 42
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Find a stable setup to run simulation tests locally #1995
Comments
Oddly I get this even after setting the tls cert timeout periods to 300 seconds, implying xud isn't starting at all, but I can't figure out why. I think at it would be good to get some error logging here to detect if xud isn't starting if possible and print any error messages. |
Building images with npm fails instantly for me, probably something wrong with my node/npm which I just installed. |
Rebooted and tried again, exact same outcome as above. |
Building and running them worked fine on my machine:
|
Looks like the problem is |
Indeed the problem is probably that |
I finally got simulation tests to build successfully with the exact same setup that worked for @michael1011 :
What I did:
Running them fails with:
|
2020-11-19T11_32_50_983Z-debug.log:
|
Should we also bump simulation tests to current versions? E.g. to go 1.14 and lnd to 0.11.1 , guess there is more to be updated. https://github.com/ExchangeUnion/xud/blob/master/test/simulation/go.mod |
Tried a couple of more runs with |
Can you set up a VM with this "default developer setup":
And harden simulation tests for this? You should be able to reproduce the issues above without problems. Our goal is to have the setup above with these specific version work reliably @LePremierHomme |
Can you try:
|
Did that, same:
|
I also had these issues. A reboot somehow solved them |
Let me try that, also good hint for @LePremierHomme |
Reboot didn't solve it for me. |
It still just looks as if you don’t have access to
Why do you run this directly? It should always be run only as part of Regarding your running failure - something is wrong with indra’s docker setup (“This node is not a swarm manager…”). I’m trying to reproduce it on gcloud VM now. |
@erkarl do you think debugging the existing Indra setup is worth it since Connext v2 is just around the corner? |
No, disable it. |
Ubuntu 20.10 has an issue with docker installation (“Package 'docker-ce' has no installation candidate”), so I used Ubuntu 20.04 LTS instead. Looks like it had no issue the swarm manager. This is what I got:
Requirements:
Setup script (some
TODO (@LePremierHomme):
|
Docker install scripts are always behind. Please install docker on ubuntu 20.10 via ubuntu repo: |
That looks like A LOT of additional dependencies that are needed. Are we sure we need all of these?
|
I thought I tried that first and it didn't work, will try it again.
These dependencies are installed due to |
Todov2:
|
And I have my first successful run. What I did:
And it went through! Full log: https://paste.ubuntu.com/p/HBZysWs4HV/ |
Second run which I started immediately after without cleaning anything succeeded too. |
Third run which I started after reboot (without cleaning or any further actions) succeeded too. |
Let's finalize the todo's @LePremierHomme and then let @sangaman do an acceptance test - he's working on xud mainly and needs to run simulation tests locally. |
I suggest to disable connext v1/indra simulation tests if they are causing issues. We'll be switching to V2 this week. |
Well they are 100% reliable for me now after #1995 (comment), I opened #2001 to close this issue for now adding hints how to prep your environment to make tests pass reliably. If I manage, everyone should ;) We'll anyways take another look at simulation tests after you did the switch to vector @erkarl |
I cleaned my environment with steps as per steps of my PR. Now the script fails to create the docker indra network again and the only solution seems to be to create it manually:
|
Following all the steps in the README in #2001 still gives me the error in the OP here:
Should we reopen? |
Well it doesn't fail for anyone else with that, thus isn't easy to debug for others. Rather open a separate issue and I guess you'll have to look into it yourself :/ Only thing that comes to mind are resources/Internet: You made sure Internet was stable and you have 10GB+ RAM free when you ran the tests? |
@sangaman are you on |
I get this every time I try to run simulation tests locally:
Attempting to run on a resource constrained (but clean/fresh) VM results in similar timeout errors.
The text was updated successfully, but these errors were encountered: