Madelyn Olson

@parthpatel Are you still planning to work on this?

Highlighting my two concerns: * How much complexity are we signing up standalone clients to implement by having them understand the concept of "redirects", that was already built for cluster...

`CLIENT TRACKING REDIRECT` I did think about that but was hoping nobody really uses is. I think basically everyone that uses server assisted client side caching uses RESP3 which doesn't...

> I can imagine some other behavior changes in the future would only make sense at the server level. Can you elaborate more, because I don't think there are that...

OK, besides the one comment about `tot-cmds`, the change LGTM.

> BTW, the PR only record the fat output buffer, we should also record the fat input buffer (by argv_len_sum). +1, although I would imagine unexpected large outputs are a...

Discussed this in a separate meeting. Everyone is generally aligned about the idea, but we need to refine the name and details.

+1 to Largelog. Alternative being considered is `HEAVYTRAFFICLOG`.

> For mocking - just to zoom into this real quick - my understanding is the process would look something like: > Write a file with a fake/mock implementation of...

@PingXie @murphyjacob4 Ok, this should be ready now. I'm a little uncertain about compatibility on all distributions though. I got some weird errors on the version of CentOS I run,...