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
S3 timing out on CI since 3.0.3 (tests using LocalStack) #1118
Comments
Apologies, it seems I previously downgraded the dependencies incorrectly. The following seems to fix the problem, after all, at least for <dependencyManagement>
<dependencies>
<dependency>
<groupId>software.amazon.awssdk</groupId>
<artifactId>bom</artifactId>
<version>2.20.63</version>
<type>pom</type>
<scope>import</scope>
</dependency>
</dependencies>
</dependencyManagement> Unfortunately, it doesn't work with
EDIT: I tried running |
It occurred to me that this is probably not the correct repo to create this issue in. I created another one for AWS SDK: While I still think that this one might be useful, feel free to close it if you don't think it's relevant for |
@daniel-frak which Localstack version are you using? It does happen that when AWS releases new SDK, Localstack sometimes stops behaving as expected. Then they release a new version and all works well. So if your issue happens still on Localstack 3.2.0, I would recommend opening an issue there. AWS team have nothing to do with LocalStack |
I tried with LocalStack |
Type: Bug
Component:
S3
Describe the bug
Running any version of
spring-cloud-aws-dependencies
higher than 3.0.2 (tested with 3.0.3 - 3.1.1) causes our Gitlab CI runner (an AWSt3.xlarge
instance) to time out on S3 operations:While the tests usually take 6-10 min, here they take up to 43 minutes before failing the build.
We are running the tests using Maven and creating a LocalStack instance using TestContainers (
1.19.7
). The relevant code is:Importantly, everything works perfectly well on a local machine (the entire build takes about 3.5min). The issues only occur on CI. Furthermore, the issue does not occur on version 3.0.2. On 3.0.2, all S3 operations take milliseconds, as expected (even on CI).
I thought that maybe it's the fault of a new version of AWS SDK, so I tried downgrading AWS SDK to the version from 3.0.2:
I found a semi-similar problem on LocalStack's Github, so I tried running LocalStack with the suggested environment variables:
None of the above fixed the problem. I tried looking through changes between 3.0.2 and 3.0.3 but I'm unable to find anything that would stand out as a likely culprit.(EDIT: It turns out AWS SDK is to blame after all. See comment below)The bucket the tests are struggling to create and then remove is called
test-bucket
, so I don't think any name-related problems are to blame.At the same time, I really need to update this dependency (I need the FIFO queue name bugfix introduced in 3.1.1) and I'm at a loss what to do.
The runner instance shows only mild memory and CPU consumption, so it doesn't seem like lack of resources is the problem here. Below is a simplified version of our
.gitlab-ci.yml
:I don't think our
ci_settings.xml
proxies are to blame, but just in case here's a simplified version of those:In summary,
some change(EDIT: the AWS SDK update) in3.0.3
is causing our CI to time out during S3 operationsand we can't figure out what it is.Sample
Unfortunately, so far I was unable to reproduce this problem outside of Gitlab CI.
The text was updated successfully, but these errors were encountered: