This is a follow up to:

https://alog.ligo-wa.caltech.edu/aLOG/index.php?callRep=42730

After a week of observing this change it does not appear to have had an impact in the retransmit rate of h1fw2.  Comparing the producer receive times from h1fw2 & h1fw1 shows that h1fw2 has a noisier signal than h1fw1.

We will probably not push this change from h1fw2 into the production system.

I have attached 3 plots.

1. The retransmits over the last 8 days.  The point of interest is the right most line, which is the retransmits since the config change.
2. A 3hr sample of the producer receive time on h1fw2 & h1fw1 for Monday 2 July 2018 (pre config change on h1fw2).
3. A 3hr sample of the producer receive time on h1fw2 & h1fw2 for Monday 9 July 2018 (post config change on h1fw2).

The h1fw2 signal looks a little cleaner than last week, but still is showing large spikes, spikes that are larger than h1fw1.  So while this may be a contributing factor to h1fw2, it appears to not be an issue on h1fw1 (which is newer hardware and cpu).