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
I'm dealing with high latency network (WAN) and the single message at a time pattern is too slow. I decided to steal the shoveling concept from RabbitMQ (basicly bulk copy messages to/from two queue systems https://www.rabbitmq.com/shovel.html ).
as soon as I did this I realized that Expiration and Visibility are stored as DateTime values. The local and remote machines are in different timezones. When I moved messages around, messages wouldn't be processed for hours.
If I make the changes to make Visibility and Expiration stored as DateTimeOffset, do you think this is worth while to bring into Rebus proper or does this not go into a direction you want to take Rebus?
The text was updated successfully, but these errors were encountered:
I guess it would not hurt to use DateTimeOffset instead of DateTime. Don't you think it would be possible to implement in a way where only newly-created tables get the DateTimeOffset type columns, while the existing send/receive SQL can work with both types?
I'm dealing with high latency network (WAN) and the single message at a time pattern is too slow. I decided to steal the shoveling concept from RabbitMQ (basicly bulk copy messages to/from two queue systems https://www.rabbitmq.com/shovel.html ).
as soon as I did this I realized that Expiration and Visibility are stored as DateTime values. The local and remote machines are in different timezones. When I moved messages around, messages wouldn't be processed for hours.
If I make the changes to make Visibility and Expiration stored as DateTimeOffset, do you think this is worth while to bring into Rebus proper or does this not go into a direction you want to take Rebus?
The text was updated successfully, but these errors were encountered: