To get on same page as to why a change was proposed in #92
Was there a timeout error occurring in playback mode before the dateTime changes:?
If so that was a bug in the worfklow steps - Playback should not have timeout workflow on.
The reasons are:
speed keep it lean and mean like backtest
event model different to realtime
checks are skipped like unsafe mode
error handling requirements are simple like backtest - not like realtime not needed
we want to keep playback mode super fast for 1 to 1000X tests
Caveats such as timeout will occur and maybe others
Solution
We do not want to change the date to fix the timeout but in fact to remove the Playback from that section .
So the solution for speed tests is not to change the date and add in the playback to the workflow
it is to fix the workflow and make sure playback does not get entangled in timeouts and so on. etc
-The dateTime Now refactor is fine #91
MicroTrendsTom
commented
1 year ago
To get on same page as to why a change was proposed in #92
Was there a timeout error occurring in playback mode before the dateTime changes:?
If so that was a bug in the worfklow steps - Playback should not have timeout workflow on.
The reasons are:
Solution We do not want to change the date to fix the timeout but in fact to remove the Playback from that section . So the solution for speed tests is not to change the date and add in the playback to the workflow
it is to fix the workflow and make sure playback does not get entangled in timeouts and so on. etc -The dateTime Now refactor is fine #91