All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=634919

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICgltC2swoM


Bot(s) for this bug's original alert(s):

win

Perf sheriff ping: reminder to follow up on possible performance issues

fdoray: I think we missed the boat on finding this in a timely manner, but is it possible that https://codereview.chromium.org/2208493002 caused the size increase? It's the only CL in the range that affects win.

Fixit ping. fdoray: any thoughts on comment 4?

Ping.

I get a 52 KB size reduction when I revert my CL. That doesn't seem enough to explain this regression https://codereview.chromium.org/2416043002. 

I built mini_installer.exe at r409923 (the suspected CL) and at r409922 (the CL just before). The size does not increase.

Branch Base Position - Mini-installer size
r409923                51 521 KB
r409922                51 531 KB

Looks like there's some kind of infrastructure issue here - there's no way either of the patches could have triggered this.

Dave, any ideas what could have caused this? Is there something specific to build-per-revision on Windows that could cause a jump?

I'm going to close this as it's old, we don't know why it happened, and sizes is being tracked at a higher level at this point.