-
Notifications
You must be signed in to change notification settings - Fork 2
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
Requesting unfiltered manifest yields a 504 #3892
Comments
@hannes-ucsc to propose browser improvement regarding large manifests taking a long time and to have Azul expose an estimated manifest creation time in the summary response. Also, @hannes-ucsc to come up with a way to accelerate the content hash check (move some of the check to indexing time). |
First @achave11 spike to eliminate a TDR slowness (listing sources) as a contributing factor. |
When looking at the above request and a few others, the following times were observed for the requests that list the TDR sources.
In the above findings, it is apparent that the content hash generation takes longest in comparison to the listing of TDR sources. |
@hannes-ucsc : "Great. That confirms that TDR is not the culprit here. Back to me for coming up with a solution to accelerate the hash generation." |
Since we're not using incremental indexing anymore, there is currently is no way the index could change without a change to the configured set of snapshots to be indexed. So for now it is safe to assume that a hash of the set of the snapshot names is a good enough approximation of the content hash. The hash should be insensitive to ordering differences, obviously. Long term, we should hash the bundle reference from all contributions into the projects index and change |
A similar situation occurred on 2024-11-13 this time with a filter that specifies all file types and species. |
When requesting a compact manifest with no filters in
prod2
a 504 Gateway Timeout response is returned.Note that the service logs for this request display no malfunction.
CloudWatch Logs Insights
region: us-east-1
log-group-names: /aws/lambda/azul-service-prod2
start-time: -3600s
end-time: 0s
query-string:
Download the manifest in cmd.exe with `curl` using:
curl.exe --location --output "hca-manifest-aa352adb-2626-503d-96ca-391ee03742ba.tsv" "https://edu-ucsc-gi-azul-dcp2-prod-storage-prod2.s3.amazonaws.com/manifests/aa352adb-2626-503d-96ca-391ee03742ba.tsv?response-content-disposition=attachment%3Bfilename%3D%22hca-manifest-aa352adb-2626-503d-96ca-391ee03742ba.tsv%22&AWSAccessKeyId=&Signature=ZuDuJAapNzDxxlDLZr0luVKJlYs%3D&x-amz-security-token=IQoJb3JpZ2luX2VjEDoaCXVzLWVhc3QtMSJHMEUCIQCITDRO7BtSSCtSHI0jVR8RElQzpGgt%2Fl4H6h9eekh2xgIgPTuwTZloKyP%2BRugCjHzO0t4pVeLemaqPOj9VAoPNQnAqoAIIk%2F%2F%2F%2F%2F%2F%2F%2F%2F%2F%2FARACGgw1NDI3NTQ1ODkzMjYiDEVQVrriymGVUos0sCr0AS5Ae1xhjhj%2FpR3IDvoiPTbePWJ76CyDkhXD6Jzo%2BBTjefBEKCQMAf2rH6qfZAbgnnx5o0fSNTPeZ81EbQ%2Be4L%2FcsGRBd4UMGfrV%2BOQyYi5mH0dS0728bgfAIBuIh1jC6EKxd5tq2pSxUde1rVoD6wFgeISDRiUhlJmjFTce2J0fQr%2FevNaOrlcN%2FO4KEBEb7oaOHHi3SgufFJfCTzo26KlBb2NNdWBlAV22GKAXaf4svPS2E9fOZDA%2BEifumdNcGO1m8GCa8FPUIgTBM3bArYHPFYpOaaJ%2BCnKrHvqjzE73v%2B27oXVTEJ6qUmTlIq
The text was updated successfully, but these errors were encountered: