conflicting results

Whilst trying to explain the many ways in which a websites speed can be improved, i used my blog as a target (a word-press instance with woeful load times and long ttfb).

Oddly, while i dramatically improved the pages speed, the results i got were conflicting.

The Dublin node http://www.webpagetest.org/result/130207_N4_KCF/ gave me the following score:

First Byte Time: A
Keep-alive Enabled: F
Compress Transfer: F
Compress Images: A
Cache static content: D
CDN detected: X

but meanwhile in Gloucester http://www.webpagetest.org/result/130207_4R_KCB/

First Byte Time: B
Keep-alive Enabled: A
Compress Transfer: A
Compress Images: A
Cache static content: C
CDN detected: X

while in London town: http://www.webpagetest.org/result/130207_T2_JS1/

First Byte Time: A
Keep-alive Enabled: A
Compress Transfer: A
Compress Images: A
Cache static content: C
CDN detected: X

So why did Dublin award the test
Keep-alive Enabled: F
Compress Transfer: F
when London and Gloucester award an A grade?

@hsiboy

Looks Gloucester and Dublin are going through PSS but Dublin isn’t so that waterfalls aren’t comparable.

Is that what you were expecting?

I re-ran the Dublin test and it came back with similar behavior as the others: http://www.webpagetest.org/result/130213_5Z_15XR/

If you had recently CNAMEd to PSS before running the test there is a good chance that your DNS changes just hadn’t propagated through the Internet yet (depending on the TTL it could be cached in various servers for hours or days).