Record the number of freshness periods since a stale HttpCache entry was last used

Record the number of freshness periods since a stale HttpCache entry was last used

This metrics records the number of freshness periods that have elapsed since a stale entry was last used. This will help us to determine the right time to prefresh resources (e.g., prefresh on cache hit if the entry's age is > 50% of freshness).

BUG=614355
Committed: https://crrev.com/91ae3602104125b4afca4f76d2b2841fda80718f
Cr-Commit-Position: refs/heads/master@{#395885}

[jkarlin, 2016-05-24 13:34:13]

jkarlin@chromium.org changed reviewers:
+ gavinp@chromium.org, rkaplow@chromium.org

[jkarlin, 2016-05-24 13:34:14]

gavinp@ PTAL at everything, thanks!
rkaplow@ PTAL at histograms.xml, thanks!

[rkaplow, 2016-05-24 17:43:26]

https://codereview.chromium.org/2005753004/diff/20001/net/http/http_cache_tra...
File net/http/http_cache_transaction.cc (right):

https://codereview.chromium.org/2005753004/diff/20001/net/http/http_cache_tra...
net/http/http_cache_transaction.cc:2805: (time_since_use * 1000) /
stale_entry_freshness_);
instead of doing this, I think you want milliseconds, right? Why not use
inmilliseconds?

https://code.google.com/p/chromium/codesearch#chromium/src/base/time/time.h&q...

https://codereview.chromium.org/2005753004/diff/20001/tools/metrics/histogram...
File tools/metrics/histograms/histograms.xml (right):

https://codereview.chromium.org/2005753004/diff/20001/tools/metrics/histogram...
tools/metrics/histograms/histograms.xml:17742: +    freshness periods have
elapsed since the entry was last used (x1000).
I'm a bit confused by the unitng (ex. the x1000 mention). You give a timedelta
which should be in microseconds - isn't this just the time delta in
milliseconds?

[jkarlin, 2016-05-24 17:57:40]

Thanks!

https://codereview.chromium.org/2005753004/diff/20001/net/http/http_cache_tra...
File net/http/http_cache_transaction.cc (right):

https://codereview.chromium.org/2005753004/diff/20001/net/http/http_cache_tra...
net/http/http_cache_transaction.cc:2805: (time_since_use * 1000) /
stale_entry_freshness_);
On 2016/05/24 17:43:25, rkaplow wrote:
> instead of doing this, I think you want milliseconds, right? Why not use
> inmilliseconds?
> 
>
https://code.google.com/p/chromium/codesearch#chromium/src/base/time/time.h&q...

I see why the 1000 is making you think of milliseconds. But the metric is
recording a ratio, not a time. The * 1000 is just to make sure the integer
division has enough precision.

https://codereview.chromium.org/2005753004/diff/20001/tools/metrics/histogram...
File tools/metrics/histograms/histograms.xml (right):

https://codereview.chromium.org/2005753004/diff/20001/tools/metrics/histogram...
tools/metrics/histograms/histograms.xml:17742: +    freshness periods have
elapsed since the entry was last used (x1000).
On 2016/05/24 17:43:25, rkaplow wrote:
> I'm a bit confused by the unitng (ex. the x1000 mention). You give a timedelta
> which should be in microseconds - isn't this just the time delta in
> milliseconds?

See the comment in the cc file.

[jkarlin, 2016-05-24 18:38:30]

https://codereview.chromium.org/2005753004/diff/20001/net/http/http_cache_tra...
File net/http/http_cache_transaction.cc (right):

https://codereview.chromium.org/2005753004/diff/20001/net/http/http_cache_tra...
net/http/http_cache_transaction.cc:2805: (time_since_use * 1000) /
stale_entry_freshness_);
On 2016/05/24 17:57:40, jkarlin wrote:
> On 2016/05/24 17:43:25, rkaplow wrote:
> > instead of doing this, I think you want milliseconds, right? Why not use
> > inmilliseconds?
> > 
> >
>
https://code.google.com/p/chromium/codesearch#chromium/src/base/time/time.h&q...
> 
> I see why the 1000 is making you think of milliseconds. But the metric is
> recording a ratio, not a time. The * 1000 is just to make sure the integer
> division has enough precision.

I've added a comment to clarify what's being recorded here.

[rkaplow, 2016-05-24 18:40:12]

lgtm

right, ok. makes sense, thanks for clarifying

[jkarlin, 2016-05-25 12:51:43]

Thanks.

gavinp@ PTAL!

[gavinp, 2016-05-25 13:52:57]

lgtm

[jkarlin, 2016-05-25 14:02:12]

The CQ bit was checked by jkarlin@chromium.org

[commit-bot: I haz the power, 2016-05-25 14:02:24]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/2005753004/40001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/2005753004/40001

[commit-bot: I haz the power, 2016-05-25 14:05:50]

Committed patchset #3 (id:40001)

[commit-bot: I haz the power, 2016-05-25 14:07:19]

Description was changed from

==========
Record the number of freshness periods since a stale HttpCache entry was last
used

This metrics records the number of freshness periods that have elapsed since a
stale entry was last used. This will help us to determine the right time to
prefresh resources (e.g., prefresh on cache hit if the entry's age is > 50% of
freshness).

BUG=614355
==========

to

==========
Record the number of freshness periods since a stale HttpCache entry was last
used

This metrics records the number of freshness periods that have elapsed since a
stale entry was last used. This will help us to determine the right time to
prefresh resources (e.g., prefresh on cache hit if the entry's age is > 50% of
freshness).

BUG=614355

Committed: https://crrev.com/91ae3602104125b4afca4f76d2b2841fda80718f
Cr-Commit-Position: refs/heads/master@{#395885}
==========

[commit-bot: I haz the power, 2016-05-25 14:07:20]

Patchset 3 (id:??) landed as
https://crrev.com/91ae3602104125b4afca4f76d2b2841fda80718f
Cr-Commit-Position: refs/heads/master@{#395885}