Is your feature request related to a problem? Please describe.
If downloading one log file takes longer than the attempt's time limit, the next attempt starts that file over. A sufficiently large or slow file may therefore never finish.
Describe the solution you'd like
Let the tool keep safe download progress between attempts. If the file changes, start that file over; never calculate a cost from a partial file. Test a download that needs several attempts, plus a file that changes midway.
Describe alternatives you've considered
A caller can increase the timeout or use smaller rolled logs, but those options may not be practical for every cluster or event log.
Additional context
Follow-up to #278. The large-event-log case came up in review: #278 (comment)
Is your feature request related to a problem? Please describe.
If downloading one log file takes longer than the attempt's time limit, the next attempt starts that file over. A sufficiently large or slow file may therefore never finish.
Describe the solution you'd like
Let the tool keep safe download progress between attempts. If the file changes, start that file over; never calculate a cost from a partial file. Test a download that needs several attempts, plus a file that changes midway.
Describe alternatives you've considered
A caller can increase the timeout or use smaller rolled logs, but those options may not be practical for every cluster or event log.
Additional context
Follow-up to #278. The large-event-log case came up in review: #278 (comment)