@Cleverbilling@_t_i_show Ah yah this isn't a 'drop a fix' sadly :( we are working to scale multiple services across Actions including our scheduler. This is 'a way off' being great to be honest, huge leaps over the last week but from pretty much 'dead to running' so still insanely far to go still
@Cleverbilling@_t_i_show The delays have got worse in the past week for some as we have been tweaking the logic we have to decide 'who goes first and who jumps where'. In parallel we have been horizontally scaling the bottleneck here. So this is somewhat 'still going' and a bit symptom of that WIP
@fardarter@moraes_c_ If you want to jump on a talk through any of these more with me -> shoot me a DM. Very happy to share more details of what is coming, where we are at, and just listen more.
Thank you again and please keep sharing
@fardarter@moraes_c_ 1. Thank you for taking the time to share so many thoughts
2. All of them would be amazing additions
3. Keep sharing -> If you care enough to share, I care enough to listen here or get on a call and talk through it more
@fardarter@moraes_c_ 12. Yep also needed. That's in the security list so that is on the cart hopefully for the next few months depending how we see the risks and attacks pivot/what energy that takes
@fardarter@moraes_c_ 9. Yeah this I need to check on
10. this would be really cool
11. aahh yes... we went deep on this a while back, this was part of our general developer quality of life work. That is slightly on ice for the moment while we do some more security bits
@fardarter@moraes_c_ 6. Again environments are just.. what they are. This makes sense. First call for us is sorting out the rest of secret scoping, then get environments on latest branch protections
7. Fair feedback, I got nothing in hand but you are right
8. Same as 7, totally fair but nothing now
@fardarter@moraes_c_ 4. Yes this + 5. are things that are on our mind. The fact that prompting is limited to environments right now is honestly mad. Both of these are things we are looking at as part of our security roadmap. Not right now, thats scoping for cache up next
@fardarter@moraes_c_ 1. Makes sense, nothing right now but I will go and poke how far we can get
2. https://t.co/vSBCrrmoG1 we are bug bashing enforce right now
3. Read/write to cache or to the repo or something else?
On August 17, GitHub experienced a significant outage that disrupted developers and organizations around the world. If you were trying to ship software that day, we let you down.
I posted in March and April about the steps we’re taking to make GitHub more reliable. The work is still underway. Here is an update:
https://t.co/9KaQy7O4CG
@mariorod1@DanielLockyer@matthewisabel I think Daniel got a DM :) but closing the loop: First set of improvements are in. For smaller log times (100-200ms ish) we are seeing around a 33% improvement. But for the real chunky ones (like up to 16 MiB) we have got these down by around 73% <3 BUT still more to do <3
1/ We went from idea to having this live insanely fast, seeing the team get creative on risks and getting something 'out the door to make life better' this fast is inspiring
2/ This is opinionated. We are taking a stance and putting in a GitHub automated process to protect people. From this side of protections for Actions this is a first for us to surface like this. Not a policy. Not a token scope. Something we detect and act upon.