Pages seem aggressively cached |
||
Issue descriptionChrome Version: (copy from chrome://version) OS: (e.g. Win7, OSX 10.9.5, etc...) What steps will reproduce the problem? (1) Use luci-milo build link (https://luci-milo.appspot.com/buildbot/tryserver.chromium.mac/mac_chromium_rel_ng/352509) (2) Refresh the page to see how much progress it's making through various test suites (3) Looks like progress is not being made (4) Check actual build link (https://build.chromium.org/p/tryserver.chromium.mac/builders/mac_chromium_rel_ng/builds/352509) and see that it's further along than what luci-milo page has It seems like the results displayed by luci-milo are either aggressively cached server-side (Cmd+Shift+R doesn't cause an update) or do not receive updates as frequently as the build links. What is the expected result? Results on luci-milo that more closely track real-time. What happens instead? Results on luci-milo seem to lag by several minutes the actual state of the build. Please use labels and text to provide additional information. For graphics-related bugs, please copy/paste the contents of the about:gpu page at the end of this report.
,
Dec 14 2016
Makes sense that it could be related to the buildbot performance issue, but I definitely was seeing a lag of more than just 30 seconds.
,
Mar 10 2017
There is caching Milo's buildbot build page, it reads the data straight from the datastore https://cs.chromium.org/chromium/infra/go/src/github.com/luci/luci-go/milo/appengine/buildbot/build.go?q=build+file:%5Einfra/go/src/github%5C.com/luci/luci-go/milo/appengine/buildbot/+package:%5Echromium$&l=518 Not sure if there ever was. Most likely there was a lag between events happening on Buildbot side and pubsub events are being sent from Buildbot to Milo Closing because there is no caching and the root cause ( bug 673202 ) is fixed.
,
Mar 10 2017
s/There is caching/There is no caching/
,
Mar 10 2017
There has been a lot of fixes put in since last Dec on reporting performance. It's expected to be about 45s behind on average, but no more than that excluding brief spikes. |
||
►
Sign in to add a comment |
||
Comment 1 by hinoka@chromium.org
, Dec 13 2016