Closed GoogleCodeExporter closed 8 years ago
We have a similar issue. And we double checked all our SDKs are up to date. The
drop is as big as 75% of usual impression count.
What we found in ours is that the overall impression to AdWhirl dropped 75%.
AdMob fill rate remained at nearly 99%.
Original comment by pinkbana...@gmail.com
on 7 Sep 2011 at 6:21
Are you seeing anything in your logs? Nothing's changed on the adwhirl end, so
if you haven't altered your config data then it shouldn't be on our side.
Original comment by wesgood...@google.com
on 7 Sep 2011 at 10:43
sorry, i know it sounds silly. but can you let me know how i can get the log?
Original comment by pinkbana...@gmail.com
on 8 Sep 2011 at 3:55
Just to clarify, we haven't made any changes on our end either, and this has
happened across all of our apps, so it must definitely be a server side issue,
or some change on that part, as otherwise I would assume this would be related
to only one or a few of our apps. And, those apps haven't been updated either
in a long time
Original comment by fred...@ticbits.com
on 8 Sep 2011 at 6:18
The problem seem to be with house ads getting weighted WAY too much for some
reason. Instead of weights like 0.5 or 0.2, all our house ads now get a weight
of 10 for some reason.
Original comment by nix...@gmail.com
on 10 Sep 2011 at 10:02
anyone still suffering from this? our impression count is still much lower than
before. we are trying to skip adwhirl and go directly with admob, will know
results in a few days.
Original comment by pinkbana...@gmail.com
on 12 Sep 2011 at 5:30
[deleted comment]
Issue 255 has been merged into this issue.
Original comment by wesgood...@google.com
on 15 Sep 2011 at 1:10
I rolled back the servers and have identified the bug (a glitch in
CustomsServlet.java's pickNid method, which was responsible for resetting
weight values in the cache, exposed because we've updated the caching layer to
save some memory).
You shouldn't be seeing this in production anymore. Tomorrow I'll see about
writing some tests to confirm my fix, and then I'll be pushing the code to our
backends.
Original comment by wesgood...@google.com
on 15 Sep 2011 at 1:14
Issue 253 has been merged into this issue.
Original comment by wesgood...@google.com
on 20 Sep 2011 at 8:55
Original comment by wesgood...@google.com
on 23 Sep 2011 at 10:54
Original issue reported on code.google.com by
fred...@ticbits.com
on 7 Sep 2011 at 8:24