Monday Dec 22, 2008

GlassFish Enterprise Server: Outstanding Price/Performance

After taking a look at Tom Daly's post on compelling price/performance, I thought I would take a slightly different approach. Tom's post covers the details of the recent 100% open source software result of 1197.1 JOPS@Standard using GlassFish, OpenSolaris and MySQL running on a Sun X4150 Server. The competitive results are also on quad-core X64 servers.

The data shown here includes 3 years of 24x7 phone support for each operating system, database, and application server.  This provides a more appropriate scenario for those deploying in a production environment. While Tom's post has the raw data, I'm just showing the updated $/JOP.

 Application Server
$/JOP

GlassFish Enterprise Server $33.90/JOP
Oracle WebLogic Server Standard Edition 10.3 $232.68/JOP
Oracle Application Server 10g Release 10.1.3.3 $193.71/JOP


Required disclosure : SPEC and SPECjAppServer are registered trademarks of Standard Performance Evaluation Corporation. Results from www.spec.org as of 11/05/2008. 2xSun Fire X4150 (8 cores, 2chips) and 1xSun Fire X4150 (4 cores, 1 chip) 1197.10 SPECjAppServer2004 JOPS@Standard; Best result with 8 cores in application-tier of the benchmark: 1xHP BL460c (8 cores,2chips) and 1xHP BL480c (8 cores,2 chips) 2056.27 SPECjAppServer2004 JOPS@Standard; Best result with 2 systems in application-tier of the benchmark: 2xDell PowerEdge 2950 (8 cores, 2 chips) and 1xDell PowerEdge R900 (4 chips, 24 cores) 4,794.33 SPECjAppServer2004 JOPS@Standard.

MySQL Pricing: https://shop.mysql.com/enterprise/
Oracle Price Book: http://www.oracle.com/corporate/pricing/eplext.pdf
Red Hat Server pricing: https://www.redhat.com/apps/store/server/
Sun OpenSolaris, GlassFish Enterprise Server pricing: http://tinyurl.com/SunSoftwareVolume


Thursday Sep 27, 2007

BEA taking notice of GlassFish

Looks like BEA is taking notice of GlassFish. Bill Roth offered some counter-arguments to our "10% better performance than BEA WebLogic 9.x" comparison, which is both expected and an appropriate thing to do. Bill asked some open-ended questions which I addressed for the benefit of the BEA community. I also offered some clarification around some of the points that Bill made so the WebLogic community has a better understanding of the GlassFish community. Bill's a busy guy (moderated comments) and the blogosphere is a moving window of conversation. In fear of missing the that window, here is a copy of the comment I submitted to Bill's blog.

Update: Alexis has a view on subject as well.
Update: Thorleif chose GlassFish over WebLogic (2nd comment) for the very reasons stated below.
Update: Hmmm, Bill never published my original comment to his blog, only the courtesy FYI ...
Update: Bill's SPAM catcher got in the way (see comments). No worries Bill, my email filter occasionally junk-mails my own emails :-)
 

---

Bill, GlassFish V2 and the Sun Java System Application Server 9.1 are essentially the same, with the latter applying the Sun brand, indemnification and support.  Sun benchmarked the Sun Java System Application Server 9.1 instead of GlassFish V2 because SPEC rules require a supported product. I highly recommend that instead of asking open-ended questions, you engage the GlassFish community via the user forums or the mailing lists where these and future questions can be easily addressed. To be completely transparent, I've updated the GlassFish User FAQ with the differences (GUI installer, 3rd party JDBC drivers, HADB) between GlassFish and the Sun Java System Application Server. These are not "significant differences" in our view.

Regarding our JVM, we don't need to be "pressed", simply asked (refer to the previous comment on user forum and mailing list). Yes, there are performance improvements between Java SE 5 and Java SE 6. However, since neither BEA WebLogic 9 nor BEA WebLogic 10 support Java SE 6, WebLogic customers cannot benefit from these performance improvements. It is good to hear that BEA has improved WebLogic tuning since the last submission since that benefits the larger Java community. The GlassFish community has made significant performance improvements between GlassFish V1 and GlassFish V2. The GlassFish community continues to actively work to improve performance across the board in a variety of areas.

I'll concede that benchmarking is a game of leapfrog. The main point is that GlassFish is performance competitive and extremely feature/price competitive at $4,500 for 4 sockets, which is why Gartner, Forrester, and Current Analysis are taking notice.

Thanks.

John Clingan
Sun Application Server Product Manager
http://blogs.sun.com/jclingan
 

About

John Clingan

Search

Archives
« April 2014
SunMonTueWedThuFriSat
  
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
   
       
Today