Bugzilla

I mean, in the context of why is this failing the regular expression matching, is this supposed to happen? Is this normal? Should the code cover this empty match continue the execution or fail it as not normal?

(In reply to Aaron Klotz [:aklotz] from comment #26)
> `Microsoft-Windows-TCPIP//win:Info` is different from the other event names that I posted in the attachment.
> 
> The regex in  `trackThreadNetIO` tries to capture the content between the slashes, but in this case the content between the slashes is empty.

I mean, in the context of why is this failing the regular expression matching, is this supposed to happen? Is this normal? Should the code cover this empty match continue the execution or fail it as not normal?


(In reply to Aaron Klotz [:aklotz] from comment #17)
> Also, keep in mind what this test is actually seeking: I/O reads and writes on the main thread are bad, especially when they occur during startup.

Considering this, the only concerning zeroes are indeed the ones provoked by the matching error during startup for [windows10-64-shippable](https://treeherder.mozilla.org/perf.html#/graphs?highlightAlerts=1&series=mozilla-central,1936478,1,1&series=mozilla-central,1936481,1,1&timerange=2592000), [windows10-64](https://treeherder.mozilla.org/perf.html#/graphs?highlightAlerts=1&series=mozilla-central,1776882,1,1&series=mozilla-central,1776885,1,1&timerange=2592000) and [windows10-64-qr](https://treeherder.mozilla.org/perf.html#/graphs?highlightAlerts=1&series=mozilla-central,1784270,1,1&series=mozilla-central,1784273,1,1&timerange=2592000). So once we get to the same page with the questions above, we can identify a soltuion.

(In reply to Aaron Klotz [:aklotz] from comment #26)
> `Microsoft-Windows-TCPIP//win:Info` is different from the other event names that I posted in the attachment.
> 
> The regex in  `trackThreadNetIO` tries to capture the content between the slashes, but in this case the content between the slashes is empty.

I mean, in the context of why is this failing the regular expression matching, is this supposed to happen? Is this normal? Should the code cover this empty match continue the execution or fail it as not normal?


(In reply to Aaron Klotz [:aklotz] from comment #17)
> Also, keep in mind what this test is actually seeking: I/O reads and writes on the main thread are bad, especially when they occur during startup.

Considering this, the only concerning zeroes are indeed the ones provoked by the matching error during startup for [windows10-64-shippable](https://treeherder.mozilla.org/perf.html#/graphs?highlightAlerts=1&series=mozilla-central,1936478,1,1&series=mozilla-central,1936481,1,1&timerange=2592000), [windows10-64](https://treeherder.mozilla.org/perf.html#/graphs?highlightAlerts=1&series=mozilla-central,1776882,1,1&series=mozilla-central,1776885,1,1&timerange=2592000) and [windows10-64-qr](https://treeherder.mozilla.org/perf.html#/graphs?highlightAlerts=1&series=mozilla-central,1784270,1,1&series=mozilla-central,1784273,1,1&timerange=2592000). So once we get to the same page with the questions above, we can identify a solution.