-
Notifications
You must be signed in to change notification settings - Fork 60
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
[ADAP-670] [Regression] communication error
in 1.5
#518
Comments
communication error
in 1.5communication error
in 1.5
Just wanted to chime in mentioning that I am seeing this error as well and can confirm that downgrading to 1.4 results in no communication error. |
Hello, EDIT: I confirm that downgrading to |
@Brooke-white, do you have any idea what might have changed in the latest patch that was released today?
|
Confirming that downgrading to |
The exact same issue on |
Hi, if anyone experiencing this issue could enable redshift-connector's logging and provide the logs it'd be a great help in the redshift-connector team's investigation of this issue. We're working to try and reproduce on our end and will cut a release as soon as a fix is identified. Thanks! |
The issue also exists on dbt-core 1.5.2 with dbt-redshift 1.5.6 which appears to install redshift-connector==2.0.912. Confirmed setting redshift-connector to 2.0.911 like others above have mentioned works to resolve this issue. Does seem to hang a lot long at the initial connection after starting a dbt task but once it starts running it works. |
@Brooke-white Here's what we saw in our CI: click to expand
|
Thanks so much, that was extremely helpful :) I reviewed the trace and was able to reproduce the issue The root cause is this line in DBT. I see the type of
|
for those interested, here is my repro:
here's my verification of fix:
happy to open a PR if you folks are ok with this fix. |
if you'll humor me being overly pedantic, this isn't the "root" cause per se. The root cause would be where in the changes issues in silly example:
|
While passing a |
At the end of the day, we don't really need to assign blame anywhere. It's just important that we solve the problem. Thank you, @Brooke-white! |
Agreed with @falonso-alo, I don't think either side (I already don't like that connotation) should be assigning blame. This industry is toxic enough as it is and this work is hard enough as it is. With that in mind, I did some research to understand exactly what is going on and felt like sharing that information. The initial confusion is coming from that fact that In version There is no check there to see whether This line was updated in with reference to this new function: The new function calls the builtin Putting all of that aside, I have to imagine we're not the only group that is passing a subclass of to this: response = socket.getaddrinfo(host=host, port=int(port), family=socket.AF_INET) Such is life with dynamic typing in python. |
Hi, seems like the error is back. I am on
I see in my docker build that |
@jan-benisek shoot! I've opened #531 to fix this once we're working on a 1.5 patch release to fix this as we speak. I'll reply back here when the patch is out, my estimate is within the next four hours, but certainly in the next 24 hours. |
Is it normal to have a stable release that doesn't include dependencies locked in? If that's even an options. 😅 |
|
@dataders Thanks a lot for such a quick fix, all 🟢 ! And don't worry, we all have been there. |
hi @dataders, i'm currently facing this blocker. do i need to downgrade the dbt-redshift adapter? |
Is this a regression in a recent version of dbt-redshift?
Current Behavior
dbt compile
fails with a "communication error"/OSError using DBT 1.5.Expected/Previous Behavior
dbt compile
succeeds using DBT 1.4Steps To Reproduce
dbt compile
Relevant log output
No response
Environment
pip packages (working)
pip packages (regression)
Additional Context
Tested in a venv.
The text was updated successfully, but these errors were encountered: