-
Notifications
You must be signed in to change notification settings - Fork 167
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
Coverity builds broken #3723
Comments
When I installed the new Jenkins workspace, I downloaded a more recent version of Coverity that I deployed on the existing machines too (and I updated the Jenkins job). This is probably the compilation going out of memory. I saw the same symptoms on Fedora hosts. The kernel kills the entire process tree when it happens, including the Jenkins agent. |
The Equinix machines (where the job was running) have more CPU/RAM (#3597 (comment)) than the IBM machine (#3597 (comment)) which I put back online a few hours ago. The job is running
which for the Equinix machine was 16 (which would appear to be 2 threads per each of the 8 cores). We could possibly set The IBM machine is 2 vCPUs/4 GB RAM, which is more like the regular test machines -- maybe adding 2GB swap like we did for the test machines would be sufficient, although the job is tending to prefer running on test-equinix-ubuntu2204-x64-1. Or maybe we can be more drastic and shift the job to the Hetzner benchmark machines? I forget if there's a reason these had to run on the |
Here's a run with hardcoded |
That build passed. I suggest to keep the hardcoded value until a better solution is implemented. |
Install the Coverity Scan build tool on the `benchmark` machines instead of the `jenkins-workspace` machines. Refs: nodejs#3723
Install the Coverity Scan build tool on the `benchmark` machines instead of the `jenkins-workspace` machines. Refs: nodejs#3723
I've updated the job to run on the benchmark machines instead of https://ci.nodejs.org/view/Node.js%20Daily/job/node-daily-coverity/3037/ looks okay apart from an expected failure to upload/submit since we're limited to one upload per day. The next scheduled daily run would be expected to pass. |
hmm. The scheduled build failed to upload: % Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0
0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0
100 183 0 0 100 183 0 139 0:00:01 0:00:01 --:--:-- 139
100 183 0 0 100 183 0 79 0:00:02 0:00:02 --:--:-- 79
100 199 100 16 100 183 4 56 0:00:04 0:00:03 0:00:01 61
100 199 100 16 100 183 4 56 0:00:04 0:00:03 0:00:01 61
error code: 1016parse error: Invalid numeric literal at line 1, column 6 Rerunning: https://ci.nodejs.org/view/Node.js%20Daily/job/node-daily-coverity/3039/console |
Do we use |
Yes, we use We are already printing the response -- in this case
I just logged into test-hetzner-ubuntu2204-x64-1 and checked the |
https://ci.nodejs.org/view/Node.js%20Daily/job/node-daily-coverity/3040/ failed 😞 % Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0
100 365 0 182 100 183 303 304 --:--:-- --:--:-- --:--:-- 608
500 Internal Server Error
If you are the administrator of this website, then please read this web application's log file and/or the web server's log file to find out what went wrong.jq: error (at response:1): Cannot index number with string "url"
parse error: Invalid numeric literal at line 1, column 13 i.e. the first call to the Coverity Scan API returned 500 Internal Server Error
If you are the administrator of this website, then please read this web application's log file and/or the web server's log file to find out what went wrong. I guess we'll need to monitor this for a while. FWIW we only have a small sample size, but the successful run was on test-hetzner-ubuntu2204-x64-2 while the two failing runs were on test-hetzner-ubuntu2204-x64-1. |
Install the Coverity Scan build tool on the `benchmark` machines instead of the `jenkins-workspace` machines. Refs: #3723
https://ci.nodejs.org/view/Node.js%20Daily/job/node-daily-coverity/3041/ succeeded. https://ci.nodejs.org/view/Node.js%20Daily/job/node-daily-coverity/3042/ looks like it succeeded on first glance, but the second stage of the upload failed 100 44 100 16 100 28 4 8 0:00:04 0:00:03 0:00:01 13
error code: 1016 https://scan.coverity.com/projects/node-js?tab=overview is currently showing "Version: v23.0.0-pre-50695e5de1" which is from 3041, but the page also says "Last Build Status: In-queue. Your build is currently being analyzed." Both builds ran on test-hetzner-ubuntu2204-x64-1. |
https://ci.nodejs.org/view/Node.js%20Daily/job/node-daily-coverity/3043/console failed to upload: Your build is already in the queue for analysis. \
Please wait till analysis finishes before uploading another build.
parse error: Invalid numeric literal at line 1, column 5 https://scan.coverity.com/projects/node-js?tab=overview: I wonder if the failed to upload build from 3042 is now blocking further uploads. I've clicked "Terminate build", which responded:
Retrying: https://ci.nodejs.org/view/Node.js%20Daily/job/node-daily-coverity/3044/ |
100 44 100 16 100 28 4 8 0:00:04 0:00:03 0:00:01 13
error code: 1016 😞 |
I logged into test-hetzner-ubuntu2204-x64-1 and manually ran the iojs@test-hetzner-ubuntu2204-x64-1:~/build/workspace/node-daily-coverity$ curl --fail-with-body -X PUT -d token=<redacted> https://scan.coverity.com/projects/<redacted>/enqueue
curl: (22) The requested URL returned error: 530
error code: 1016 I immediately ran it again and it succeeded: iojs@test-hetzner-ubuntu2204-x64-1:~/build/workspace/node-daily-coverity$ curl --fail-with-body -X PUT -d token=<redacted> https://scan.coverity.com/projects/<redacted>/enqueue
{"project_id":6507,"id":619487} (I've added
|
Small correction, the upload is a three step process:
Of the observed failures so far:
|
https://ci.nodejs.org/view/Node.js%20Daily/job/node-daily-coverity/3045/console failed at step 1: 100 199 100 16 100 183 4 53 0:00:04 0:00:03 0:00:01 57
07:49:03 curl: (22) The requested URL returned error: 530
07:49:03 error code: 1016 |
https://ci.nodejs.org/view/Node.js%20Daily/job/node-daily-coverity/3046 failed at step 3: 100 44 100 16 100 28 4 8 0:00:04 0:00:03 0:00:01 13
07:53:23 curl: (22) The requested URL returned error: 530
07:53:23 error code: 1016 I've manually run step 3 on the machine to unstick the analysis queue in Coverity. I'll put a loop around the first and third steps so it retries a few times (with a pause between attempts). |
Typically the three most recent Coverity builds since I added the retry loops all succeeded without having to retry 😆. |
Since I put in the retry loop, we've only had one build failure, which occurred during the build (possibly a resource issue or agent failure): https://ci.nodejs.org/view/Node.js%20Daily/job/node-daily-coverity/3056/ All other builds have succeeded and were able to submit the results to Coverity without needing to go through the retry loop, so we have no validation that the loop works/makes things better. Since the builds are succeeding at the moment and we're getting the static analysis run daily I'm going to close this issue. |
Two most recent node-daily-coverity builds have failed.
Some error about the agent going offline, no obvious other error.
e.g. https://ci.nodejs.org/view/Node.js%20Daily/job/node-daily-coverity/3010/console
The text was updated successfully, but these errors were encountered: