Re: macosx 1 process, linux N processes, macosx chokes under load

This WebDNA talk-list message is from

2001


It keeps the original formatting.
numero = 36846
interpreted = N
texte = >On 6/28/01 3:38 PM, The Defendant Dale LaFountain >Confessed: > >>> On 6/28/01 12:34 PM, The Defendant Dale LaFountain >>> Confessed: >>> >>>> >>>> OSX is NOT always slower. Running a single instance of my template, >>>> OSX is faster by 50%. It's only when multiple requests occur >>>> simultaneously that OSX crawls, even in comparison to OS9/Webstar on >>>> the same hardware. >>>> >>>> Isn't it possible that some inefficiencies exist in the way >>>> webcat/osx handles threads that could cause this massive difference >>>> in performance? Do other applications that run on both OSX and Linux >>>> have this problem, or is this a problem specific to webcat? If not >>>> in webcat, then the problem should be reported to the Darwin group >>>> and to Apple as a bug. >>>> >>>> If this is an inherent flaw in (current) OSX, I would strongly >>>> discourage anyone from using webcat/osx for busy webcat-driven sites. >>>> It certainly can't keep up with our (relatively simple) testing suite >>>> using Netsplat. >>>> >>>> -Dale >>> >>> Dale, What hardware are you using? >>> >>> I just did a new install on a 533 DP. I will have a gig and a half of ram >>> in it by next week. If you like, I'm more than willing to run your test >>> suite and see how it compares to linux/ or single proc machine. >>> >>> >>> -- Aaron Lynch >> >> We're testing on a dual 533 G4 as well, with 512 mb ram. >> >> You're welcome to test, but I'm guessing you'll see the same results. >> >> -Dale > >Thanks for the info. What are you using on the Linux side? >I am not in any way needing to get every ounce of performance out of this >particular box. It will be pretty low volume. But I'm curious just the >same. It will be interesting to see if the situation changes as X matures. > > >-- Aaron LynchIt's a dual P3-550 running Redhat 7.0 and a software mirrored disk. With the performance increase in 10.0.4, I'm hopeful that this performance gap will resolve itself soon (fingers crossed).> >Can you post the tests you did - I would be interested to run them on my OSX >Server 1.2 box to compare speeds. Will post findings back to the list. > >Also - can you define 'load' to the list. > >Thanks >AlainWe define load as 5 or more simultaneous requests, as requested by the Netsplat traffic simulation app.We put together a test suite that covered several areas: cpu intensive no databases, db intensive searches, and db intensive replace/appends. Each of these tests is configurable in intensity (1-3 nested searches, 10-10000 results), all configurable in the url so as to run different sets of templates through Netsplat.We also just took a variety of pages from our live e-commerce site (www.tfaw.com) and starting hitting them with 10 splats at a time. The performance sucks on OSX when the number of splats is above 2, but is acceptable on OS9. I'm setting up an even faster linux box shortly (dual p3-933) to do more testing.I'm also working on a summary of our findings, which covers OS9, OS10 Server 2.0, and Linux versions of webcat. I'll post a link to an archive of our test suites in a couple days.-Dale-- Dale LaFountain Vice President of Information Technology, WebMaster Dark Horse Comics, Inc. mailto:dalel@darkhorse.com Phone: 503-652-8815 x324 http://www.darkhorse.com/------------------------------------------------------------- This message is sent to you because you are subscribed to the mailing list . To unsubscribe, E-mail to: To switch to the DIGEST mode, E-mail to Web Archive of this list is at: http://search.smithmicro.com/ Associated Messages, from the most recent to the oldest:

    
  1. Re: macosx 1 process, linux N processes, macosx chokes under load (Aaron Lynch 2001)
  2. Re: macosx 1 process, linux N processes, macosx chokes under load (Dale LaFountain 2001)
  3. Re: macosx 1 process, linux N processes, macosx chokes under load (Alain Russell 2001)
  4. Re: macosx 1 process, linux N processes, macosx chokes under load (Aaron Lynch 2001)
  5. Re: macosx 1 process, linux N processes, macosx chokes under load (Dale LaFountain 2001)
  6. Re: macosx 1 process, linux N processes, macosx chokes under load (Aaron Lynch 2001)
  7. Re: macosx 1 process, linux N processes, macosx chokes under load (Alain Russell 2001)
  8. Re: macosx 1 process, linux N processes, macosx chokes under load (Dale LaFountain 2001)
  9. Re: macosx 1 process, linux N processes, macosx chokes under load (Le Pham 2001)
  10. macosx 1 process, linux N processes, macosx chokes under load (Dale LaFountain 2001)
>On 6/28/01 3:38 PM, The Defendant Dale LaFountain >Confessed: > >>> On 6/28/01 12:34 PM, The Defendant Dale LaFountain >>> Confessed: >>> >>>> >>>> OSX is NOT always slower. Running a single instance of my template, >>>> OSX is faster by 50%. It's only when multiple requests occur >>>> simultaneously that OSX crawls, even in comparison to OS9/Webstar on >>>> the same hardware. >>>> >>>> Isn't it possible that some inefficiencies exist in the way >>>> webcat/osx handles threads that could cause this massive difference >>>> in performance? Do other applications that run on both OSX and Linux >>>> have this problem, or is this a problem specific to webcat? If not >>>> in webcat, then the problem should be reported to the Darwin group >>>> and to Apple as a bug. >>>> >>>> If this is an inherent flaw in (current) OSX, I would strongly >>>> discourage anyone from using webcat/osx for busy webcat-driven sites. >>>> It certainly can't keep up with our (relatively simple) testing suite >>>> using Netsplat. >>>> >>>> -Dale >>> >>> Dale, What hardware are you using? >>> >>> I just did a new install on a 533 DP. I will have a gig and a half of ram >>> in it by next week. If you like, I'm more than willing to run your test >>> suite and see how it compares to linux/ or single proc machine. >>> >>> >>> -- Aaron Lynch >> >> We're testing on a dual 533 G4 as well, with 512 mb ram. >> >> You're welcome to test, but I'm guessing you'll see the same results. >> >> -Dale > >Thanks for the info. What are you using on the Linux side? >I am not in any way needing to get every ounce of performance out of this >particular box. It will be pretty low volume. But I'm curious just the >same. It will be interesting to see if the situation changes as X matures. > > >-- Aaron LynchIt's a dual P3-550 running Redhat 7.0 and a software mirrored disk. With the performance increase in 10.0.4, I'm hopeful that this performance gap will resolve itself soon (fingers crossed).> >Can you post the tests you did - I would be interested to run them on my OSX >Server 1.2 box to compare speeds. Will post findings back to the list. > >Also - can you define 'load' to the list. > >Thanks >AlainWe define load as 5 or more simultaneous requests, as requested by the Netsplat traffic simulation app.We put together a test suite that covered several areas: cpu intensive no databases, db intensive searches, and db intensive replace/appends. Each of these tests is configurable in intensity (1-3 nested searches, 10-10000 results), all configurable in the url so as to run different sets of templates through Netsplat.We also just took a variety of pages from our live e-commerce site (www.tfaw.com) and starting hitting them with 10 splats at a time. The performance sucks on OSX when the number of splats is above 2, but is acceptable on OS9. I'm setting up an even faster linux box shortly (dual p3-933) to do more testing.I'm also working on a summary of our findings, which covers OS9, OS10 Server 2.0, and Linux versions of webcat. I'll post a link to an archive of our test suites in a couple days.-Dale-- Dale LaFountain Vice President of Information Technology, WebMaster Dark Horse Comics, Inc. mailto:dalel@darkhorse.com Phone: 503-652-8815 x324 http://www.darkhorse.com/------------------------------------------------------------- This message is sent to you because you are subscribed to the mailing list . To unsubscribe, E-mail to: To switch to the DIGEST mode, E-mail to Web Archive of this list is at: http://search.smithmicro.com/ Dale LaFountain

DOWNLOAD WEBDNA NOW!

Top Articles:

Talk List

The WebDNA community talk-list is the best place to get some help: several hundred extremely proficient programmers with an excellent knowledge of WebDNA and an excellent spirit will deliver all the tips and tricks you can imagine...

Related Readings:

BBEdit/HTMLcomments/WebCat/[/FONT] (1999) Problems getting parameters passed into email. (1997) Summing fields (1997) Server Freeze (1998) WebCat2b12 CGI Mac - [shownext] problem (1997) emailer on Windows Beta 18 (1997) Pricing (1998) Categories (2000) WebDNA-generated email headers (2003) Shownext on AOL (2003) PCS Frames (1997) WebCatalog and WebMerchant reviewed by InfoWorld (1997) [OT] (waaaay OT) (2004) stripping extra spaces from form fields (2000) ShipCost Problems (1998) Not really WebCat - need HTML Grider (1997) apostrophe in search item (1997) Next (1997) Strange intermittent WebDNA problems (2008) SSL, WebSTAR, WebCatalog (1998)