Bedrock Computer Technologies, LLC v. Softlayer Technologies, Inc. et al

Filing 667

Opposed MOTION to Strike Untimely Supplemental Expert Report of Dr. Mark Jones by AOL Inc, Amazon.com Inc., Google Inc., MySpace Inc., Softlayer Technologies, Inc., Yahoo! Inc.. (Attachments: #1 Declaration of T. Briggs, #2 Exhibit A - Email from A. Hendershot, #3 Exhibit B - Supplemental Report of Jones, #4 Text of Proposed Order)(Jones, Michael)

Download PDF
Bedrock Computer Technologies, LLC v. Softlayer Technologies, Inc. et al Doc. 667 Att. 3 Exhibit B Dockets.Justia.com IN THE UNITED STATES DISTRICT COURT FOR THE EASTERN DISTRICT OF TEXAS TYLER DIVISION BEDROCK COMPUTER TECHNOLOGIES LLC, Plaintiff, v. SOFTLAYER TECHNOLOGIES, INC., et al. Defendants. CASE NO. 6:09-cv-269 Jury Trial Demanded SUPPLEMENTAL REPORT: SUPPLEMENTAL TEST RESULTS I submit this report as a supplement to my infringement report. Specifically, the Defendants and their experts have lodged a number of criticisms regarding my methodology of the tests I presented in my Opening Expert Report. When I served that report, I also produced the modified source code that I tested, the data measured by my tests, my testing protocol, the code for one of the programs that I used to generate traffic, and the specifications of the hardware I was testing. In my opinion, I provided the Defendants and their experts with everything they would need to replicate my tests. In this Supplemental Test Results, I used the same testing methodology to address the Defendants' criticisms and further to quantify and verify my expert report and deposition testimony. I incorporate the tests and discussion of those tests from my Opening Expert Report into this supplement. The purpose of my tests was to compare the performance results of the modified code with and without the call to subroutine rt_free() associated with the rt_genid deletion in rt_intern_hash(). My testing methodology in these tests is consistent with my original tests. For this test, I used the same parameters and conditions as used in Appendix P of my Opening Report. For the runs without rt_free(), I used the same code. For the runs with rt_free(), I used the code produced in my Opening Report with the single change that I removed the "/* */" that commented out the call to rt_free(). These results demonstrate that, for approximately 80% of the runs, the code without the call to rt_free() runs more efficiently. This is consistent with my earlier results, and this is why I disabled the call to rt_free() in my original tests. Further, this confirms my analysis that calling rt_free() results in additional work that adversely affects performance. While these tests show that the Defendants' criticisms of my tests are incorrect, this exercise has -2- uncovered a mistake I made in Appendix P of my Opening Report. Specifically, in the test that generated the data for the condition "Cache Enabled" and "'120 Disabled," the memory on the server was running low. This was likely leading to these cases to have slower performance. This mistake is isolated to the data in Appendix P, and I did not form my opinion of the 11-25% performance improvement from the '120 patent based on these results. The cause of this mistake was that the amount of memory consumed by the routing table was significantly larger in the tests for Appendix P than the tests associated with the other Appendices. The routing cache table associated with Appendix P was sixteen times the size of the routing table in the tests associated with Appendices H and N and four times the size of the routing table in the tests associated with Appendix O. Moreover, I verified that this mistake was isolated to Appendix P by re-examining the remainder of the results in the tests in Appendices H, N, O and P to verify that there was sufficient, available memory. Executed on March 16, 2011 ________________________ Mark T. Jones -3- Average Number of URL Average Successes Number Per Second IP IP of URL CPU Scaled by Hping Address Address Successes CPU Busy Interrupt CPU Busy Interval Repeat Working No Per Processing Processing Processing (sec) Count Set rt_free rt_free Second Percentage Percentage Percentage 64 16 4 0 64 16 4 0 64 16 4 0 64 16 4 0 64 16 4 0 64 16 4 0 64 16 4 0 64 16 4 0 64 16 4 0 64 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 40 40 40 40 40 40 40 40 40 40 40 40 40 100000 100000 100000 100000 100000 100000 100000 100000 150000 150000 150000 150000 150000 150000 150000 150000 250000 250000 250000 250000 250000 250000 250000 250000 100000 100000 100000 100000 100000 100000 100000 100000 150000 150000 150000 150000 150000 1 1 1 1 0 0 0 0 1 1 1 1 0 0 0 0 1 1 1 1 0 0 0 0 1 1 1 1 0 0 0 0 1 1 1 1 0 0 0 0 0 1 1 1 1 0 0 0 0 1 1 1 1 0 0 0 0 1 1 1 1 0 0 0 0 1 1 1 1 0 0 0 0 1 2156 2018 2004 2018 2171 2033 2014 2014 2163 2027 1994 1994 2167 2056 2034 2020 2119 1969 1959 1858 2172 2027 2021 2023 2146 2047 2010 2019 2166 2027 2010 2009 2164 2010 2008 2016 2169 23.5 25.2 28.3 30.6 24.4 25.8 29.9 31.4 21.6 23.8 32.5 31.6 25.1 27.5 30.9 33.9 25.2 30.3 34.1 40.1 26.1 32.1 39.3 44.7 23.4 25.0 28.1 31.0 23.6 25.4 27.9 30.2 23.0 25.7 30.4 31.0 24.0 9.3 12.2 15.4 17.8 10.1 12.9 16.8 18.4 8.7 12.4 18.8 19.0 11.3 14.0 17.9 21.4 10.9 17.0 20.6 27.5 12.5 19.1 26.3 31.3 9.3 11.7 15.1 17.8 9.4 11.9 15.4 17.5 9.5 12.0 16.1 18.0 10.3 9161 8022 7075 6603 8891 7880 6745 6418 10023 8514 6143 6301 8619 7474 6582 5955 8402 6499 5749 4635 8315 6305 5138 4529 9169 8178 7165 6508 9187 7974 7200 6658 9404 7832 6608 6493 9045 Percentage Difference Between URL Scaled Results for the "rt_free" vs "no rt_free" 2.9 1.8 4.7 2.8 14.0 12.2 -7.1 5.5 1.0 3.0 10.6 2.3 -0.2 2.5 -0.5 -2.3 3.8 16 4 0 64 16 4 0 64 16 4 0 40 40 40 40 40 40 40 40 40 40 40 150000 150000 150000 250000 250000 250000 250000 250000 250000 250000 250000 0 0 0 1 1 1 1 0 0 0 0 1 1 1 0 0 0 0 1 1 1 1 2053 2040 2043 2098 1945 1936 1886 2167 2028 2023 2034 26.7 30.3 33.2 24.4 30.0 34.6 39.6 26.9 32.1 38.2 44.3 13.0 17.1 19.9 10.8 16.8 22.1 26.7 12.8 19.4 25.3 31.0 7678 6726 6149 8584 6477 5593 4759 8054 6313 5290 4588 2.0 -1.8 5.3 6.2 2.5 5.4 3.6

Disclaimer: Justia Dockets & Filings provides public litigation records from the federal appellate and district courts. These filings and docket sheets should not be considered findings of fact or liability, nor do they necessarily reflect the view of Justia.


Why Is My Information Online?