Henri van de Scheur - Sun Norway wrote:
Kathey Marsden wrote:
As an aside, it would be wonderful to hear from some of the folks
that make up the "We" in Henri's mails that say "We have run [snip
lots of great testing]" about their opinions and votes on the release.
I am so sorry!
'We' are:
Ole Solberg
Vemund Østgaard
John Embretsen
Henri van de Scheur
[snip]
Details for our additional test are found under
http://www.multinet.no/%7Esolberg/public/Apache/10.1.3.1_RC/10.1.3.1RC_Additional_Tests.html
(you will see the link to this page after you do the refresh/reload?)
As some of you may have noticed, "we" did not run a (so-called) long
running test against 10.1.3.1.
However, I have run a test against 10.1.3.0-RC (as noted in [1]) for
about 13 days, 18 hours. This is the same test that caused
OutOfMemoryErrors against the 10.1.2.4 snapshot after about 21 days [2].
Unfortunately, due to lack of time and some other constraints (e.g.
physical relocation of test hosts), the 10.1.3.0-RC test was not (and
most likely will not be) run long enough to see the full effect of the
anti-Memory-leak medicine provided by DERBY-1103.
However, graphs showing Network Server heap (tenured space) utilization
in 10.1.2.4 vs. 10.1.3.0-RC (for as long as the latter test ran) are
attached to this e-mail. DERBY-1103 seems to have had a positive (as in
good) effect on memory usage.
See attachments
10.1.2.4_OldSpaceUsage_first350hrs.png and
10.1.3.0-RC_OldSpaceUsage.png.
--
John
[1]
http://www.nabble.com/Re%3A-Sun-DBTG-testing-for-10.1.3.0---Results-p4974557.html
[2] http://www.nabble.com/Re%3A-Testing-for-10.1.3-p4688077.html
