All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=922523

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=349b58a133371c4b5ccaf805db9e6d47ef3012247f34d02b23a9b7a77647708d


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

Google Chrome Win

grt: there is a regression of about 500k in chrome.dll.pdb. As the test owner, can you determine how important it is to fix? Do you see anything from the commit log that looks like a likely culprit?

https://chromium.googlesource.com/chromium/src/+log/ab9292fa9d43591613adb3871ff2d39672216369%5E..fe595425bcc114d73dc8a8fd6fbca738cc78e0b8?pretty=fuller&n=1000

zturner: we have a growing pile of these alerts for variability in the size of the various pdbs. is it expected that the pdb size jumps around a lot?

many of the alerts i see in this query appear to be related to the pdb sizes:

https://bugs.chromium.org/p/chromium/issues/list?can=1&q=summary:%22regression%20in%20sizes%22&sort=-id&colspec=ID%20Pri%20M%20Stars%20ReleaseBlock%20Component%20Status%20Owner%20Summary%20OS%20Modified

should we change the thresholds so that these don't fire so much? are these worthy of investigation?

thanks for your help.

I filed bug 924302 on the alerts, I think we're monitoring too many things.