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
STORY: As an IMAGE scaling developer, I want to know how many simultaneous requests pegasus can sustain on an ongoing basis, so that we have an idea how much a single server can potentially scale
#18
Open
jeffbl opened this issue
Sep 19, 2023
· 2 comments
jeffbl
changed the title
As an IMAGE scaling developer, I want to know how many simultaneous requests pegasus can sustain on an ongoing basis, so that we have an idea how much a single server can potentially scale
STORY: As an IMAGE scaling developer, I want to know how many simultaneous requests pegasus can sustain on an ongoing basis, so that we have an idea how much a single server can potentially scale
Sep 19, 2023
jeffbl
transferred this issue from Shared-Reality-Lab/IMAGE-server
Oct 5, 2023
Status at end of Oct02 sprint: Created initial loadtest bash script to use testset.py from the IMAGE-testing repo to make requests at regular intervals, to see what breaks. Next steps for Oct23 sprint:
figure out what actual errors are happening, and what they mean (e.g., watch logs, check request responses)
integrate further into existing test scripts, e.g., for selecting subset of photos to use in test, instead of explicitly listing by number
Status at end of Oct23 sprint: Clear that to do comparisons of correct resultsand what comes back from heavily loaded server will require using existing test scripts, since there will be trivial differences (e.g., timestamps) that will make a naive compare fail even when results are correct. Teasing apart the scripts to work well for this testing will require more time. Importance of server scaling is reduced for upcoming grant, so moving to backlog until this becomes more pressing.
No description provided.
The text was updated successfully, but these errors were encountered: