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
Problem: To simulate certain failure scenarios, it is important to have control over when network failure happens. For example, the application layer sends message A and B in succession, and we want the network partition to kick in after A is sent but before B is sent on a TCP connection. Equivalent on UDP maybe something like when A and B are sent, drop A but allow B.
Currently it is hard to simulate those kinds of failures. It would be great to add a support to take a predicate that is run on message stream and when the predicate returns true, a fault is introduced.
For example, following would introduce a partition between host 1 and 2 after a particular message is sent from host1 to host2
Problem: To simulate certain failure scenarios, it is important to have control over when network failure happens. For example, the application layer sends message A and B in succession, and we want the network partition to kick in after A is sent but before B is sent on a TCP connection. Equivalent on UDP maybe something like when A and B are sent, drop A but allow B.
Currently it is hard to simulate those kinds of failures. It would be great to add a support to take a predicate that is run on message stream and when the predicate returns
true
, a fault is introduced.For example, following would introduce a partition between host 1 and 2 after a particular message is sent from host1 to host2
The text was updated successfully, but these errors were encountered: