You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
boto3 supports passing in a max_attempts key to this dictionary which sets the maximum number of request retries. The default value is 3 but we'd like to increase this value because we are seeing flakey behavior from some object storage providers.
[Optional] Implementation
You could allow users to set the envvar S3_MAX_ATTEMPTS.
The text was updated successfully, but these errors were encountered:
Hey @ricor07, as mentioned in the comment to the attached PR, you can already specify the download_retry argument to StreamingDataset to get download retry functionality across any cloud provider.
@jamin-chen going to close out this issue as per our discussion on #853
🚀 Feature Request
Currently the retries in
S3Downloader
are configured like so:streaming/streaming/base/storage/download.py
Lines 249 to 251 in a18648b
boto3 supports passing in a
max_attempts
key to this dictionary which sets the maximum number of request retries. The default value is 3 but we'd like to increase this value because we are seeing flakey behavior from some object storage providers.[Optional] Implementation
You could allow users to set the envvar
S3_MAX_ATTEMPTS
.The text was updated successfully, but these errors were encountered: