Forum OpenACS Q&A: Uptime Server Types [OT]

Collapse
1: Uptime Server Types [OT]
Posted by Mat Kovach on 02/25/03 03:30 AM
With the recent .NET vs OpenACS discussion that was brought up I went through the servers available on Uptime (http://uptime.openacs.org/) and pulled all the sites that hand been checked in the past 24 hours and ran a quick check on what servers were being monitored by uptime, the results are available here:

http://shell1.alal.net/~kovachme/uptime-server-types.txt

(I grouped, Apache, Microsoft-IIS, AOLserver, Zope, adn IBM_HTTP).

The top of the list is:

Server-Type Returned from Machine              Total
( ERROR Server down,
  N/A nothing returned)
-------------------------------------------    ------
Apache                                      1623
N/A                                        840
Microsoft-IIS                              566
ERROR                                      128
AOLserver                                  101
Zope                                        30
Netscape                                    12
Zeus/4.1                                    7
Rapidsite/Apa/1.3.27                        6
Stronghold                                  5
Squeegit/1.2.5                              5
NetWare-Enterprise-Web-Server/5.1          5
Zeus/3.4                                    4
WebSite/3.1.13                              4
WebSTAR/4.5(SSL)                            4
NaviServer/2.0                              4
Jetty/4.2.4rc0                              4
IBM_HTTP                                    4

(128 sites were down when I ran the test).

Thought people might find the results the interesting.  According to Uptime, AOLserver is the thrid most popular returned server type on the 'net.

A few interesting ones that I seen were:

Oracle                                      1
(Oracle?!)

OSU/3.10a;Multinet                          1
(VAX VMS)

FileMakerPro/5.5v2                          1
FileMakerPro/5.0                            1
AppleShareIP/6.3.3                          1
AppleShareIP/6.3.1                          1
(Hmm, Apple?)

Collapse
2: Re: Uptime Server Types [OT] (response to 1)
Posted by Brian Fenton on 02/25/03 10:37 AM
Wow! Only another 27 to go before we catch those ERROR guys! 😉
Collapse
3: Re: Uptime Server Types [OT] (response to 1)
Posted by Vadim Makarov on 02/25/03 10:25 PM
Fun stats. They are heavily biased towards AOLserver because Uptime service itself is a part of OpenACS community.

For more objective results, you've got to refer to something like recent NetCraft survey. It's a bit further down the list there.

Collapse
4: Re: Uptime Server Types [OT] (response to 1)
Posted by Vadim Makarov on 02/26/03 04:01 AM
The best would be to find a survey encompassing only the sites with functuionality OpenACS allows to build. I'm not sure there is one. Netcraft probably doesn't do justice to AOLserver, because it may be used for more-complex-than-average projects while Apache and IIS power a lot of small and/or static sites.
Collapse
5: Re: Uptime Server Types [OT] (response to 4)
Posted by Mat Kovach on 02/26/03 06:32 AM
I will say I did the stats to start just for fun.  While there might be a slight bias towards AOLserver because of the association between OpenACS and Uptime (and the relative small sample of 3400 websites) I would have to say that the numbers are rather reflective of the current state of the 'net.

Remember AOLserver is used for more that just OpenACS.  Just in running the uptime server and talking with some of the people that use the services, many of them have never heard of OpenACS (something I wish to change with some more linking back to OpenACS from uptime).

If you look at the usage states (http://uptime.openacs.org/usage/ and yes, I know that haven't been runnig lately) qutie a few of the referrers actually some from searchs or links from web hosting companies.  Without much "marketing" from uptime or OpenACS there is quite a bit of activity from people who really just want to monitor their website.

I think if OpenACS and AOLserver actually had information in docs that said 'Use uptime' that the numbers might be more biased, but since that is not really done (-and I am not say ing that it should be done-just saying that is isn't) I think the numbers reflect something on the 'net, I'm just not exactly sure what or who the represent.

Maybe I should take these numbers and compare them to somthing like Netcraft to see how the percentage of /monitored/ servers reflect against the percentage of servers.  Then we might be able to say, Server X has a higher percentage of /cluefull/ admins :->