Re: NetSplat and WebCat2

This WebDNA talk-list message is from

1997


It keeps the original formatting.
numero = 11662
interpreted = N
texte = >Hi all, > >I've been setting up my server(PwrMac 7100/66) and WebCat2 site and >decided to try some simulated traffic with Maxum's NetSplat. >NetSplat is a small application that can simulate web client loads of >more than 2500 connections per minute. (Like I'm going to get that >many.) >I set the client load to 60 with random delays of no more than 1 second >between requests. I set WebSTAR (it has 50mbs of RAM assigned to it) to >a max 25 connections. The page requests in the test, included one large >GIF, one small, one straight HTML page, and one .tmpl page that >generates a new cart. These are randomly requested. >I started the simulation and the server churned along great. >Then I logged on and started to cruise the site during the test. >It delivered pages a little slower than I had been seeing but not much. >Then I hit the WebCat2 search .tmpl pages, I could see the server >completely stop serving the other simulated requests. I assume, the CPU >was dedicated solely to WebCat at that point. Then when the search was >complete, the server resumed sending pages. >Back on my browser I awaited the search results. What I got was a >partial page. I tried my search again, and got another partial page. >I went back to WebSTAR and lowered the max connections to 9 and ran the >same NetSplat configuration with much more success but still >occasionally got a partial page from a search request. > >My questions, for those who have gone before, is this... > >Is NetSplat a worthy indicator of high traffic problems? >Is 60 clients requesting pages every half second or so heavy traffic? >Is this a fair test for WebCat searches? >And why the partial pages? Why not just a server busy try back later? > > >-- >Eddie Schrieffer >The Creative Edgethen grant said....>Can you try this test with the plugin instead of the CGI and see if >there's a difference? Thanks! > >Grant Hulbert, V.P. Engineering | ===== Tools for WebWarriors ===== >Pacific Coast Software | WebCatalog Pro, WebCommerce Solution >11770 Bernardo Plaza Court | SiteEdit Pro, SiteCheck, PhotoMaster >San Diego, CA 92128 | SiteGuard >619/675-1106 Fax: 619/675-0372 | http://www.smithmicro.comWe also use a 7100/66 webstar 2.0.2...we find the following performance relative to webcat2.0f3 not using netsplat. MacTCP2.0.6 beats OT whatever hands down on this type of Mac (Nubus...we have 2 other 6100/66 webservers that run macTCP because OT was way slower) If the search/results template has a lot...I mean a lot to display (see http://www.panavise.com/comm/indash/idbrand.html) the .acgi responds better then the plug-in. The plug-in works much faster if the search/results template is smaller in the quantity of what it has to display. Now will the final release of WebCat Gold plug-in vs .acgi be equall? The is something Grant would have to tell you relative to stack space for plug-ins and memory allocation, etc. Today we are running our site on the .acgi. We have been switching back and forth between the plug-in and the .acgi about every 5 to 7 days to compare how people feel about our site and it's response time.I would be curious also if you are using the webstar cache plug-in for images and if you are running all .html through webcat, or just .tmpl's?PS: If you think webcatalog is slow from your test....think about using FMpro3.0 and whatever to interface with WebStar....:)=============================================== Gary Richter PanaVise Products, Inc. 7540 Colbert Dr. Reno, Nevada 89511 Ph: 702.850.2900 Fx: 702.850.2929 Email: grichter@panavise.com http://www.panavise.com =============================================== Associated Messages, from the most recent to the oldest:

    
  1. Re: NetSplat and WebCat2 (bob 1997)
  2. Re: NetSplat and WebCat2 (grichter@panavise.com (Gary Richter) 1997)
  3. Re: NetSplat and WebCat2 (Grant Hulbert 1997)
  4. NetSplat and WebCat2 (Eddie Schrieffer 1997)
>Hi all, > >I've been setting up my server(PwrMac 7100/66) and WebCat2 site and >decided to try some simulated traffic with Maxum's NetSplat. >NetSplat is a small application that can simulate web client loads of >more than 2500 connections per minute. (Like I'm going to get that >many.) >I set the client load to 60 with random delays of no more than 1 second >between requests. I set WebSTAR (it has 50mbs of RAM assigned to it) to >a max 25 connections. The page requests in the test, included one large >GIF, one small, one straight HTML page, and one .tmpl page that >generates a new cart. These are randomly requested. >I started the simulation and the server churned along great. >Then I logged on and started to cruise the site during the test. >It delivered pages a little slower than I had been seeing but not much. >Then I hit the WebCat2 search .tmpl pages, I could see the server >completely stop serving the other simulated requests. I assume, the CPU >was dedicated solely to WebCat at that point. Then when the search was >complete, the server resumed sending pages. >Back on my browser I awaited the search results. What I got was a >partial page. I tried my search again, and got another partial page. >I went back to WebSTAR and lowered the max connections to 9 and ran the >same NetSplat configuration with much more success but still >occasionally got a partial page from a search request. > >My questions, for those who have gone before, is this... > >Is NetSplat a worthy indicator of high traffic problems? >Is 60 clients requesting pages every half second or so heavy traffic? >Is this a fair test for WebCat searches? >And why the partial pages? Why not just a server busy try back later? > > >-- >Eddie Schrieffer >The Creative Edgethen grant said....>Can you try this test with the plugin instead of the CGI and see if >there's a difference? Thanks! > >Grant Hulbert, V.P. Engineering | ===== Tools for WebWarriors ===== >Pacific Coast Software | WebCatalog Pro, WebCommerce Solution >11770 Bernardo Plaza Court | SiteEdit Pro, SiteCheck, PhotoMaster >San Diego, CA 92128 | SiteGuard >619/675-1106 Fax: 619/675-0372 | http://www.smithmicro.comWe also use a 7100/66 webstar 2.0.2...we find the following performance relative to webcat2.0f3 not using netsplat. MacTCP2.0.6 beats OT whatever hands down on this type of Mac (Nubus...we have 2 other 6100/66 webservers that run macTCP because OT was way slower) If the search/results template has a lot...I mean a lot to display (see http://www.panavise.com/comm/indash/idbrand.html) the .acgi responds better then the plug-in. The plug-in works much faster if the search/results template is smaller in the quantity of what it has to display. Now will the final release of WebCat Gold plug-in vs .acgi be equall? The is something Grant would have to tell you relative to stack space for plug-ins and memory allocation, etc. Today we are running our site on the .acgi. We have been switching back and forth between the plug-in and the .acgi about every 5 to 7 days to compare how people feel about our site and it's response time.I would be curious also if you are using the webstar cache plug-in for images and if you are running all .html through webcat, or just .tmpl's?PS: If you think webcatalog is slow from your test....think about using FMpro3.0 and whatever to interface with WebStar....:)=============================================== Gary Richter PanaVise Products, Inc. 7540 Colbert Dr. Reno, Nevada 89511 Ph: 702.850.2900 Fx: 702.850.2929 Email: grichter@panavise.com http://www.panavise.com =============================================== grichter@panavise.com (Gary Richter)

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:

About Include (2001) Search/sort in URL Was: GuestBook example (1997) Re:Searching for ALL / empty form field (1997) WCS Newbie question (1997) WebCatalog 2.0 b 15 mac (1997) would this be a good use of [spawn]? (2002) Loop into field (1998) Crashing Server (2000) WebCatalog [FoundItems] Problem - LONG - (1997) Re:listfiles-looking for slick solution (1997) Max Record length restated as maybe bug (1997) WebMerchant when CC network is down (1998) Ok here is a question? (1997) Cart Question (1998) [WebDNA] Raspberry Pi (2013) testing testing 1.2.3 (2004) [shownext max (1997) I give up!! (1997) Date and Comparisons (2002) Automatic POST arg sending? (1998)