Will Microsoft IIS Overtake Apache? 303
First time accepted submitter jcdr writes "February's 2014 Web Server Survey by Netcraft shows a massive increase [in the share of] Microsoft's web server since 2013. Microsoft's market share is now only 5.4 percentage points lower than Apache's, which is the closest it has ever been. If recent trends continue, Microsoft could overtake Apache within the next few months, ending Apache's 17+ year reign as the most common web server."
why not? (Score:4, Funny)
Re: why not? (Score:5, Informative)
If I remember correctly, Microsoft was paying large hosting providers like GoDaddy to use IIS over apache
Re: why not? (Score:5, Funny)
Next: Paying consumers to use Surface instead of iPad as their go-to breakdancing training device.
NETCRAFT CONFIRMS IT! (Score:3, Funny)
NETCRAFT IS DEAD!
Re: (Score:2)
... in soviet Russia!
Re: (Score:3)
Re: (Score:2)
Re: (Score:3)
Re:NETCRAFT CONFIRMS IT! (Score:5, Funny)
Natalie Portman's agent just woke up in a cold sweat in fear that she is no longer held dear by ./ers
But thankfully this post was made, so Natalie Portman's agent can roll over and go back to sleep.
Just not with Natalie Portman.
Because I'm doing that right now.
Giggety.
Re: (Score:2)
Yeah, we were all watching you on youtube thanks to some anonymous hacker and your laptop's webcam
IIS is Isis's sis (Score:3, Funny)
That Apache had to save
From her hirsute marital bliss
With abundant
Burma Shave
Re: why not? (Score:5, Informative)
Re: why not? (Score:5, Insightful)
"I don't know if Microsoft paid them, but GoDaddy did move all of their parked sites to IIS by default instead of Apache, which caused a major percentage change for Microsoft."
And why not, especially if Microsoft is paying them to do it? Those parked sites only represent a miniscule fraction of bandwidth, but as you say, make a big percentage difference in perceived market share.
Smooth move, Microsoft. You bring "lying with statistics" to a whole new level.
Re: why not? (Score:5, Interesting)
If I remember correctly, Microsoft was paying large hosting providers like GoDaddy to use IIS over apache
The evidence of that is the "all sites" graph which shows IIS's share increasing vs. the "active sites" graph which shows IIS's share plummeting. IIS appears to be hosting a lot of dead sites, ironically.
Re: why not? (Score:5, Funny)
IIS appears to be hosting a lot of dead sites
Which is good news for the IIS performance metrics MS will be releasing... :-)
Re: why not? (Score:5, Insightful)
The evidence of that is the "all sites" graph which shows IIS's share increasing vs. the "active sites" graph which shows IIS's share plummeting.
I think the most interesting graph is the last: 1 million busiest sites. The downtick of Apache looks a lot like the opposite of the uptick for nginx. For busy sites, it seems nginx is separating from Google and IIS, but at the expense of Apache.
Re: why not? (Score:5, Interesting)
At this point, I'm not 100% sure what in any reasonable configuration Apache would offer over nginx.
A couple of things i've noticed
1: The combination of nginx and php can be a pain. It's easy enough to make it work for the root of a hostname but if you then add a subdirectory of the domain that is mapped to a different local directory it breaks because nginx passes the wrong path to php. I belive it's possible to make things work again with a sufficiantly complex configuration but I haven't figured out how yet. In my case I just worked arround it by using subdomains.
2: Some more specialist stuff may rely on specific apache modules that afaict don't have an nginx equivilent. For example mod_dav_svn or mod_mirrorbrain.
Re: (Score:3)
Nginx doesn't do dynamic content by itself so when you have dynamic content on a site hosted on nginx it has to pass those to another process over a suitable protocol (http, fastcgi, uwsgi etc).
502 bad gateway usually indicates a problem with that "other process", that may be caused by administrative actions , bugs in the backend process or overloading but either way it isn't really nginx's fault. You can get similar errors from apache or squid, you just don't see them deployed in this manner as often.
Re: why not? (Score:5, Informative)
mmm, the "active sites" graph looks far more stable, apache is showing a slight downward trend recently but the market share it's losing doesn't seem to be going to MS
Re: (Score:2)
It's pretty standard for retail products. Why would it be different for software?
Re: (Score:3, Funny)
Re: why not? (Score:4, Interesting)
> Windows can't run a month w/o needing a reboot
Regardless, I reboot all my linux, unix and bsd boxes once a month too. Just to be sure anything can rebooted by a noc monkey without disturbing my sleep.
During daytime, yes. Redundancy is a lie unless you test it periodically.
Re: (Score:3)
Re:why not? (Score:5, Interesting)
Re: why not? (Score:2, Informative)
SQL-injections usually result in an output of data or a valid login session or destruction of data. Remote code execution would be a rare thing.
Bots usually prefer the taste of client computers. Massive botnets written in PHP inhabiting LINUX servers? Sure there are plenty of bugs in PHP, but that has nothing to do with either Apache or LINUX.
Your point is invalid, my hair is a bird!
Re: (Score:3, Funny)
The reason for knee jerk reactions is probably because the article actually shows no notable uptick in Microsoft's market share of active sites. It's just a sensationalist summary of some poorly analyzed data. For doing actual web serving (not just parked domain serving), they've fallen to 3rd, being beaten by both Apache and nginx. According to the numbers, 93.0% of Microsoft's sites are inactive, and they are leveraging 86.1% of the growth in inactive sites. Microsoft is now the leading web server for
Re: (Score:3, Funny)
Rubbish. I get that it's supposed to be humorous but usually one would base humor on some kernel of fact and there is none in your post.
Isn't ntoskrnl.exe working for you ?
It was bound to happen (Score:3, Insightful)
Apache is turning into one of the dinosaurs of the information age, being overtaken by the likes of Nginx and Lighttpd left and right but refusing to die already. IIS also is hardly the crippled pile of steaming crap which it used to be.
Re:It was bound to happen (Score:5, Funny)
IIS also is hardly the crippled pile of steaming crap which it used to be.
This is very true. It's made a lot of progress in the past few years, and is now an almost unrecognizable, completely new pile of steaming crap.
Re: (Score:3)
You mean they put a Ribbon on it? That seems to be Microsoft's go-to technique for refreshing crap heaps.
large hosting company using IIS != IIS popularity (Score:5, Insightful)
Netcraft says, "Microsoft gained a staggering 48 million sites this month, increasing its total by 19% â" most of this growth is attributable to new sites hosted by Nobis Technology Group." I have no idea WFT Nobis Technology Group is, but that suggests that what is essentially one large installation swings Netcraft's idea of "the most common web server."
And that's a broken way of counting. If ten servers using Server A serve ten sites each, and one server with Server B serves 1,000 sites,Server A is still the most common web server, with ten times the installation base of Server B.
Re:large hosting company using IIS != IIS populari (Score:5, Informative)
Exactly. A bit of sensationalism in the story.
All Sites (included millions of parked) are in 38%/%32 mix. Looking 600 pixels down and you see the active (non parked sites). The percentage is 52% vs 11%. The big drop in for MS in 2009 was probably a nail in the coffin...
Re: (Score:3)
A hosting company that mainly hosts spam sites. Together with their parent company they have large swathes of bots on various small /26 IP ranges registered to them which seems intended to be to prevent other companies from easily blocking a large IP range.
This and more (Score:5, Insightful)
The "active" sites shows no such growth trend, in fact it shows IIS declining. NginX is the only web server showing growth, and even this is misleading. Most of our use for NginX is does not make Apache go away. We use NginX as a front end reverse proxy that talks to Apache back ends. NginX is good at a few things, but nowhere near as robust as Apache.
This is just another case of pulling only the statistics you want to color a lie.
Re:large hosting company using IIS != IIS populari (Score:5, Insightful)
If 10,000 Web sites are served from one server using Apache, and 100 Web sites are served from 100 servers using IIS, it would be reasonable to interpret that Apache is the more common choice for serving Web sites. It would be reasonable--not necessarily accurate, but in a vacuum decision there is a great chance of validity--to assume that Apache is the better choice for hosting Web sites in most cases, as it has been selected for more often. It would be very reasonable to assume that Apache is, in most cases, at least adequate--a satisfiser would find this palatable--while making no assumptions on whether it is more or less optimal than IIS.
It's silly to assume that the number of servers has any real meaning, unless it can reflect resource use--at our resolution we can't even do that (are these 100 IIS servers run from Raspberry Pi, or 100 IIS servers run from ginormous Dell R620s? How much load?). Even then, that doesn't reflect all the other decisions put into it. On the other hand, there are very real questions like "Does my ASP.NET site run better on Apache?" and the answer is no; or like, "Does my Python/cherrypi site run better through WSGI/Apache or WSGI/IIS?" and the answer is no again.
The raw number of Web sites run on Apache reflects a lot more than the number of discrete servers. But then you have questions like: are these Perl/PHP/Python, .NET, etc.? Essentially: are they Apache/IIS sites because of Apache/IIS, or because of the system that provides facilities for the site best also providing Apache/IIS support best?
Re: (Score:3)
From Nobis Web site (sounds dodgy just from the language):
"Nobis Technology Group, LLC is the parent holding company to roughly a dozen specialized companies and a broad spectrum of websites. We are privately-held, employee-owned, and have been involved in a number of very lucrative Internet services companies of many names since 2002."
I don't trust NoScript to let them further out of the box for their alleged web site to tell me more.
Very different when ... (Score:5, Interesting)
Developer January 2014 Percent February 2014 Percent Change
Apache 98,129,017 54.50% 94,741,928 52.68% -1.81
nginx 21,548,550 11.97% 24,206,737 13.46% 1.49
Microsoft 20,901,626 11.61% 21,196,966 11.79% 0.18
Google 15,386,518 8.54% 15,245,912 8.48% -0.07
Re:Very different when ... (Score:5, Funny)
It also looks very different if you sort them by name:
Apache
Google
Microsoft
nginx
Re:Very different when ... (Score:5, Funny)
Much more different if you sort the words by letter.
aacehp
eggloo
cfimoorst
inngx
nginx is replacing apache (Score:2)
Re: (Score:2)
Sensationalist summary :( (Score:5, Informative)
One needs to look beyond the first graph that shows all sites surveyed to look at the actually active sites - there Apache appears to have more *active* deployments than the rest combined. Counting inactive, parked domains is not really indicative of particular server popularity.
Gee, look at how full that parking lot is (Score:5, Insightful)
Re: (Score:2)
Also, due to the new ICANN email verification requirement, [gandibar.net] there is going to be an increase in the number of "parked" domains.
From TFA (Score:5, Insightful)
I'm starting to believe the hearsay: Slashdot has really been totally overrun by astroturfers (in this case paid by Microsoft). Maybe dice sells a number of "promotional posts" on a biased article to various companies, one of them being Microsoft?
Re: (Score:2)
Maybe dice sells a number of "promotional posts" on a biased article to various companies, one of them being Microsoft?
Or maybe they're trying a "two wrongs make a right" balance for the garbage "IE's share is plummeting!!!" post from last week?
Re: (Score:2, Interesting)
Hi, I am the submitter of this story. Believe or not, I am really not connected to Microsoft, directly or indirectly. I am the co-founder of a very small company that provides services for embedded electronics and almost exclusively use Linux. I submitted this story after reading the monthly Netcraft email news. I wondering how Microsoft marketing team could possibly use the potential news "ISS > Apache" and how the OSS community will react to that news.
Reading many comments, It look like the "ISS > A
Re: (Score:3)
Sorry, but I am not connected either nor to Dice, not to Microsoft. I submitted the story simply because I wanted to understand the ground of the fast ISS raise in the first graph.
Statistics? (Score:4, Interesting)
This is a count of sites running web services, right? Not volume served out by each brand of server.
Microsoft has had the practice of starting IIS on practically every server for the purpose of providing a web management interface. In some cases, without informing the system admin.
Anecdote:
Many years ago, when I managed a few Intranet sites at Boeing (SunOS, HP-UX, AIX, Linux), we had a variant of the Code Red [wikipedia.org] worm infecting IIS systems. Admins of *NIX systems could see the propagation of the worm payload in our web logs, even though our systems were immune*. We collected the source IPs of infected systems and turned them over to computing security. The next thing we knew, we'd get calls from Windows server admins, claiming that their systems could not be infected, as they were not running IIS. "Look again." Configuring many services automatically triggers a start of IIS. And now you've got a service running that the admins don't know that they have to keep patched. So even when Microsoft released a fix, it never got applied since many admins figured it wasn't applicable to them. I would venture a guess that most Windows Server (and many client) systems are running IIS, even if it only displays the default installation page.
*Typical Apache/*NIX systems just replied with a 404 since the target DLL didn't exist. But I wrote a Perl CGI that would capture the query source and fire back a Windows popup message to the effect that their machine was broadcasting an infection. I was surprised to see how many people with client (desktop) systems called me to ask when was going on.
Re: (Score:2)
If you're not intending to run a web server then you shouldn't even have one installed...
And if something is installed, it needs to be patched even if it isn't running.
Re: (Score:2)
Don't tell me. Tell Microsoft.
Either silently starting IIS is a ploy to boost numbers. Or they figure MCSEs are too dense to read prerequisites*, so they just do stuff without asking/telling.
*This might be unfair. Its possible that they did something and got a Mr. Clippy popup that informed them of the step. But they just selected "Do it" and forgot. So, yeah. Dense.
Re: (Score:2)
A fresh install of server 2000, 2003, 2008 or 2012 server does not install IIS.
On the other hand some software will activate some IIS features but not the OS installation.
How about sum total of OSS web servers (Score:3)
Nginx http://www.nginx.org/ [nginx.org] ( really popular and at least this is in one of the graphs)
Lighttpd http://www.lighttpd.net/ [lighttpd.net] (personally, I have found many reasons use this one in the past and I'm sure I will again)
Cherokee http://www.cherokee-project.co... [cherokee-project.com] (yet to explore past a basic setup)
Roxen Webserver http://www.roxen.com/products/... [roxen.com] (Still need to take for a spin)
And then special purpose web servers.
HTTP Explorer http://http-explorer.sourcefor... [sourceforge.net]
HFS HTTP File Server http://www.rejetto.com/hfs/ [rejetto.com]
At least that's all I can think of. Anybody else?
I know some of these take up negligible market share, but I would still like to see their market share lumped together.
Re: (Score:2)
In the world of Open Source, I would also like to see the sum total of Open Source web servers VS. IIS
Why? What makes "Open Source" so magical?
IIS is actually pretty nice (Score:3)
I recently became acquainted with it at work, and it's actually quite nice to work with, I must say.
Still, this post reminds me quite a lot of the xkcd about extrapolating off of one data point. It seems unlikely that IIS will overtake Apache; more likely there was a one-time shift due to some particular event.
Huh (Score:2)
Looks like obfuscation to me, and so not accurate.
Oblig (Score:5, Funny)
At the time Elvis Presley died in 1977, he had 150 impersonators in the US. Now, according to calculations I spotted in a Sunday newspaper colour supplement recently, there are 85,000. Intriguingly, that means one in every 3,400 Americans is an Elvis impersonator. More disturbingly, if Elvis impersonators continue multiplying at the same rate, they will account for a third of the worldâ(TM)s population by 2019.
http://crookedtimber.org/2005/... [crookedtimber.org]
Re: (Score:2)
Simple answer: (Score:2)
So finally (Score:2)
The report from Netcraft: (Score:5, Funny)
It is official; Netcraft now confirms: Apache is dying
One more crippling bombshell hit the already beleaguered Apache community when IDC confirmed that Apache market share has dropped yet again, now down to less than a fraction of 1 percent of all servers. Coming close on the heels of a recent Netcraft survey which plainly states that Apache has lost more market share, this news serves to reinforce what we've known all along. Apache is collapsing in complete disarray, as fittingly exemplified by failing dead last in the recent Sys Admin comprehensive networking test.
You don't need to be a Kreskin to predict Apache's future. The hand writing is on the wall: Apachefaces a bleak future. In fact there won't be any future at all for Apache because Apache is dying. Things are looking very bad for Apache. As many of us are already aware, Apache continues to lose market share. Red ink flows like a river of blood.
All major surveys show that Apache has steadily declined in market share. Apache is very sick and its long term survival prospects are very dim. If Apache is to survive at all it will be among web server dilettante dabblers. Apache continues to decay. Nothing short of a cockeyed miracle could save Apache from its fate at this point in time. For all practical purposes, Apache is dead.
Fact: Apache is dying
*ring* *ring* (Score:2)
In part it's lies, here's a true story (Score:5, Interesting)
Re: (Score:2, Troll)
you're high
Re: (Score:3)
Why has Apache started to lose ground?
Re:IIS better in almost every way. (Score:5, Insightful)
This kind of thing happens on a regular basis and is usually due to Microsoft making backroom deals with operators of parked domains, probably not paying in cash but in Windows license discounts for servers or hosting. Borderline illegal and classic Microsoft - don't ever be fooled into thinking that Microsoft has gotten itself a corporate personality transplant. The active sites graph tells the real story: Microsoft continues to languish. It is beyond me why Microsoft is so fixated on manipulating Netcraft stats.
Re: (Score:2)
Re: (Score:2)
Probably because it costs them exactly $0 to do it.
Really? What about the team of thugs assigned to do the job, and associated hangers on?
Re: (Score:3)
... It is beyond me why Microsoft is so fixated on manipulating Netcraft stats.
They're attempting to exploit our herd mentality in order to hide their weakness; if enough of us can be fooled into thinking that IIS is more popular than it actually is, then more people will switch to it or stick with it for that reason alone.
Re: (Score:2)
Some real brilliant math here. Apache often runs on windows. You understand that the licensing costs and TCO of operating systems vs webservers is slightly different, yes ?
Re: (Score:3, Interesting)
I for one switched all my web servers from apache to nginx. It simply performs better.
Re: (Score:2)
I for one switched all my web servers from apache to nginx. It simply performs better.
If tech news has taught me anything, it's that assessing performance is never simple. I have no doubts that whatever site you migrated performs better, but that probably was the result of a single important factor to you.
Re: (Score:2)
Because it's incredibly easy to pad numbers and claim them as fact without focusing on the real details?
Re: (Score:3)
Why has Apache started to lose ground?
nginx & lighttpd
Re:IIS better in almost every way. (Score:4, Informative)
Re:IIS better in almost every way. (Score:5, Informative)
you most likely are going to be using Apache/Nginx.
IIS market share dipped a bit after 2008 and is now back to about where it was. Apache jumped a lot since 2008 and is now back a bit below where it was. Nginx has gone from 1% to 14% in the same time. IIS has hovered between 20-30% for a while. It's now closer to 30%. Apache has been in the 50-70% range for a long time, but is now dipping a lot. The only reason we're using Apache is that Nginx doesn't work as a reverse SSL proxy in front of Jenkins (apparently it can, with some magic incantations, but they didn't work for us). For everything else, Nginx is an obvious choice. It's somewhat sad to see that Nginx has completely displaced Lighttpd, as it would have been nice to have some more active competition.
Re:IIS better in almost every way. (Score:5, Interesting)
Re: (Score:3)
Re:IIS better in almost every way. (Score:4, Informative)
I have used both, from working for a hosting provider as an administrator, to working for MS itself later, to my job now. Configuration of apache is only slightly more difficult in that you need to use a text editor instead of point and click hold handing, however IIS is no wear near as powerful in that you can do much more with apache because of the slightly more difficult configuration. In addition apache is more lightweight, and has better security.
It got a little confusing about which it you were referring, so I FTFY. Why is it that after I've written the comment I get options to Submit - Continue Editing - Preview - Cancel , but not Login?
Re: (Score:2)
Re: (Score:3, Insightful)
Is IIS easier than Apache to configure in Linux?
Re: (Score:3, Funny)
Re: (Score:3, Informative)
I use both, EVERY SINGLE DAY
IIS changes how it's managed with every single point release, and nothing is obvious about it.
Just to use SSL you need to use multiple administration interfaces to import the cert and then assign that to a specific server
If you know what you are doing, and you should, you can much more easily navigate a text file with all the options than you can clicking around in a dozen places to work on IIS configs.
Then there's all the issues with file permissions
Re: (Score:2, Insightful)
constantly in need of restart and quickest to get owned by crackers
it's rubbish, bad enough to have to fight with it for internal use but only an idiot would expose that to internet. yes, had to deal with IIS for over a decade
Re:IIS better in almost every way. (Score:5, Insightful)
Considering that the grow was caused because some big parked domains (with static pages) moved to IIS, i'd say that by a very wide margin, the main use of IIS is to serve domains with just one static page.
Regarding the "better in almost every way", is almost as funny as the article title.
Re:Probably (Score:5, Insightful)
Sounds more like you just hate the industry you work in. It's probably best that you're leaving.
Re: (Score:3)
Sounds like my colleauge. "I hate Javascript" "I hate tablets" "I hate Micsosoft". "I hate PDF files" "I hate non-vi editors"
For all X where X is not Gentoo Linux and VI he hates X
Everytime he hates I just know since he hates X he has never worked with X so he in incompetent when it comes to X. And have always been correct in that assumption.
Re: (Score:2)
*ring* *ring* *ring*
You hear that? That is the 1990s calling. They want their nostalgia back. (I say this as someone who used to be a CNE and who started his career with Netware)
Do you work for the government, either Federal, State or local? That is the only place I see Novell anymore.
Re:Probably (Score:5, Insightful)
Banks, they're extremely conservative and NDS's ability to replicate a sub-portion of the directory to each branch location helps keep bandwidth usage down, which can be important if you have hundreds or thousands of locations in podunk towns. I can also see using it if you're a anti-MS shop as it's the best directory server other than AD.
Re:Probably (Score:4, Insightful)
Novell totally blew it. It was a sad day as I watched NT 4.0 servers creep into our environment. Novell had LDAP based NDS long before Microsoft cobbled together AD. It was a much better solution and they brought it to market way ahead of the competition.
Re:Probably (Score:4)
"pass the hash" and "mimikatz"... two serious problems with AD...
Re: (Score:3, Informative)
Re: (Score:2)
I predict a long and fruitful career, lolzers.
now that I'm retiring
Or not...
Re: (Score:2)
Re:I'm switching to IIS! (Score:5, Funny)
The MS shills are out in force posting as AC, you mean?
Re: (Score:2)
Yet you are such a breathtaking, marvelous specimen of humanity that you would grace this undesirable place to enlighten the muck-dwelling Unixites. Truly we are all blessed. Let this day go forth! From this day forward, every February 4th shall be known as -
Oh, wait. You didn't even bother attaching a name to this pathetic holier-than-thou screed. Back to the comforting bosom of Mommy and Windows forums with you, you pathetic bastard.
Re:I'm switching to IIS! (Score:4, Funny)
You lot remind me of my dad. All he listens to is Pink Floyd and other hippie music, he's convinced any music after 1980 is shit, so he doesn't even listen to any of it.
You should listen to your dad.
Re: (Score:2)
Any headline which ends in a question mark can be answered by the word no.
This is not true. There is probably some statistical probability which says that most headlines which end in a question mark can be answered with a "no". But not all of them.
Re: (Score:2)
Of course this one is also relevant.
http://xkcd.com/1022/ [xkcd.com]
Re: (Score:2)
You offer no facts as to why one is better than the other, other than "me and my friends think so".
Look, if you want to talk popularity, talk desktop share, use real numbers. Then explain why something with 1/100 the popularity is even in the conversation.
Otherwise admit that either one has some points over the other, neither is perfect, and admit that 'number of installs' is probably about as meaningless