Re: OT: performance FUD
On Wed, Nov 30, 2005 at 02:58:00PM -0600, William A. Rowe, Jr. wrote: fork() is *painfully* slow on the darwin kernel, I haven't tested but can't imagine that threading isn't a huge win here. Explain? One preforked worker process can handle thousands of requests. Apache doesn't have to fork for each one.
Re: OT: performance FUD
I'm referring to shrinking or growing the pool of threads/processes as needed. If worker grows threads as needed, or even has to spawn only one more process to create dozens of threads, this is goodness. Bill Brian Candler wrote: On Wed, Nov 30, 2005 at 02:58:00PM -0600, William A. Rowe, Jr. wrote: fork() is *painfully* slow on the darwin kernel, I haven't tested but can't imagine that threading isn't a huge win here. Explain? One preforked worker process can handle thousands of requests. Apache doesn't have to fork for each one. .
Re: OT: performance FUD
On Thu, Dec 01, 2005 at 04:02:49AM -0600, William A. Rowe, Jr. wrote: I'm referring to shrinking or growing the pool of threads/processes as needed. If worker grows threads as needed, or even has to spawn only one more process to create dozens of threads, this is goodness. But is it a huge win (your words)? What ratio do you observe for number of forks in relation to number of incoming requests? With processes being pre-forked ahead of demand, how large a problem does this pose in practice?
Re: OT: performance FUD
Nick Kew wrote: On Tuesday 29 November 2005 20:49, Justin Erenkrantz wrote: --On November 29, 2005 3:40:11 PM -0500 Paul A Houle [EMAIL PROTECTED] wrote: * prefork and worker seem to be about equally fast on linux? Yup - this is because forking and threading are equivalent (by and large) on Linux. That's the conventional wisdom for static stuff. If DBD catches on it'll change. Also, 2.2 proxy stuff works much better with worker in Linux. Anything that uses pooling (apr_reslist) seems to work better under worker. This may be because the fallback (ie, no threads) in most modules is kind of nasty. -- Brian Akins Lead Systems Engineer CNN Internet Technologies
Re: OT: performance FUD
Nick Kew wrote: Hmmm, how about an early adopters page? We could *imply* the organisations by featuring mugshots and brief profiles of both Brian and Colm as having successfully beta-tested 2.1.x in very-high-volume production environments. Perhaps, as long as it wasn't tied directly to turner/cnn. -- Brian Akins Lead Systems Engineer CNN Internet Technologies
Re: OT: performance FUD
Colm MacCarthaigh wrote: Well, lets not even come close to risking that. I'm in the process of having an internal whitepaper I did being sanitized for public consumption. Once that's done -- hopefully this week or next -- everything in it is quotable. -- Brian Akins Lead Systems Engineer CNN Internet Technologies
Re: OT: performance FUD
Jess Holle wrote: So if one uses worker and few processes (i.e. lots of threads per), then Solaris should be fine? That's what people think, but I'd like to see some numbers. I've never put a worker Apache into production because most of our systems depend on PHP or something else which I wouldn't trust 100% in a threaded configuration. Now that I think about it, there is a common situation where people with modest web sites (at the 50,000 ranking in Alexa) have performance problems with Apache... That's the case of people doing downloads of big (1 M files.) Conventional benchmarking, which fetishizes a large and constant number of connections on a LAN doesn't model the situation well (it doesn't model any real-world situation well.) The trouble you have a population of people with really bad connections that take forever to download things... Back when I had dialup, I used to download ISO images, I'd just use a download manager and have my computer running overnight to do it. For one project I work on, we have people uploading files that sometimes are in the ~1 M range, then we do processing on the files that is sometimees extensive. We were worried that some processes were running for 20, 30, 40 minutes, but we discovered that many of our users have horrible connections. The result is that a site with a modest number of hits per day can have 1000 simultaneous connections. With prefork you end up burning a lot more RAM than really seems fair -- although it's not so bad if you can afford to load your machine with 8G.
Re: OT: performance FUD
On Wed, Nov 30, 2005 at 11:01:55AM -0500, Paul A Houle wrote: So if one uses worker and few processes (i.e. lots of threads per), then Solaris should be fine? That's what people think, but I'd like to see some numbers. I've never put a worker Apache into production because most of our systems depend on PHP or something else which I wouldn't trust 100% in a threaded configuration. Is there anything we can do in 2.4/3.0 that will help gain that trust? -- Colm MacCárthaighPublic Key: [EMAIL PROTECTED]
Re: OT: performance FUD
Colm MacCarthaigh wrote: I've never put a worker Apache into production because most of our systems depend on PHP or something else which I wouldn't trust 100% in a threaded configuration. Is there anything we can do in 2.4/3.0 that will help gain that trust? PHP, or it's extensions or whatever they call them, are not thread safe. So until that's fixed, nothing we can do. Probably the same with other stuff. worker wokrs for me, though. -- Brian Akins Lead Systems Engineer CNN Internet Technologies
Re: OT: performance FUD
Colm MacCarthaigh wrote: On Wed, Nov 30, 2005 at 08:01:41AM -0500, Brian Akins wrote: Colm MacCarthaigh wrote: It's public knowledge that CNN.com runs Apache 2.2, would it cause you a lot of trouble for that to be referenced? As long as it's from public sources (netcraft, HTTP headers, etc) and no one within the orginization is quoted, it shouldn't be a problem, I suppose. I would be hesitant to say yes -- it's my job on-the-line if somebody here doesn't like it :( Well, lets not even come close to risking that. I have been given word that our statements can appear in a press release or testimonial, it just has to be passed through legal here. So what type of statements are we looking for? -- Brian Akins Lead Systems Engineer CNN Internet Technologies
Press release for httpd 2.2 (was Re: OT: performance FUD)
Brian Akins wrote: I have been given word that our statements can appear in a press release or testimonial, it just has to be passed through legal here. So what type of statements are we looking for? Let's bring the Apache Public Relations Committee into this to see what advice they have. We (the Apache HTTP Server Project) would like to put together a press release announcing the release of version 2.2. We can highlight the obvious market penetration numbers and some of the features listed here: http://httpd.apache.org/docs/2.1/new_features_2_2.html But we would like to put particular emphasis on the performance and scalability of httpd. We have two people (Brian Atkins at CNN and Colm MacCárthaigh at HEAnet) who run httpd under extremely high load and are willing to give us some quotes on this topic. Can the PRC give advice on how we should proceed? I am willing to start a draft, if that helps. Is there a template or some advice on format/content? What else should we be doing? Re Brian's question above, I think we want a statement emphasizing performance and scalability. It doesn't need to be extremely precise, spec-wise. Most people reading a press release wouldn't care. We just want to transmit the message that you use apache httpd in a very-high load situation and that it perfoms very well. By the way, are you using 2.0 or 2.1? Joshua.
Re: Press release for httpd 2.2 (was Re: OT: performance FUD)
Joshua Slive wrote: By the way, are you using 2.0 or 2.1? BOth. 2.1 in higher traffic now. -- Brian Akins Lead Systems Engineer CNN Internet Technologies
Re: OT: performance FUD
Colm MacCarthaigh wrote: Is there anything we can do in 2.4/3.0 that will help gain that trust? It's not Apache's fault. It's not even PHP's fault. It's a much bigger problem with the open source libraries that people link into PHP, Perl, Python and the like. The problem is particularly perceived as a PHP problem because (1) PHP is the market leader, and (2) the PHP developers are a lot more responsible than, say, the Python developers, who tell you to go ahead and write threaded apps in Python anyway. I suppose that the PHP developers could set up some system where extensions are marked as being threadsafe or not, and there's a lock on every untrusted module, then do a program of certifying modules as safe, but that's a ~big~ project: race conditions and deadlocks are a bitch to debug, particularly when the problems are in somebody else's code. PHP's market position is as a product that any idiot can download and install, just following the instructions, and get a system with good reliability and performance -- a painful phase of shaking out threading bugs would endanger that perception. The best thing I can see Apache doing is some kind of hybrid model which works like mod_event for static pages, and passes off (some or all) dynamic requests to something like prefork. Dynamic requests would eat more memory than worker, but you don't have the problem of using a heavyweight mod_perl or mod_php process spending two hours blasting bits out of a file to somebody on dialup. A process-based model is always going to be more reliable than a thread-based model. A hand grenade can go off in an server process, a server process can hemorage memory terribly, and nobody gets hurt. The user on the other end just hits 'reload' and goes on hs way.
Re: OT: performance FUD
Brian Akins writes: Colm MacCarthaigh wrote: I've never put a worker Apache into production because most of our systems depend on PHP or something else which I wouldn't trust 100% in a threaded configuration. Is there anything we can do in 2.4/3.0 that will help gain that trust? PHP, or it's extensions or whatever they call them, are not thread safe. So until that's fixed, nothing we can do. Probably the same with other stuff. worker wokrs for me, though. -- Brian Akins Lead Systems Engineer CNN Internet Technologies Exactly. I run a website for a former employer which still gets considerable traffic in terms of downloads of large files and runs a medium-to-heavy use forum system based on PHP. Multiple sites built with the LAMP stack (P being PHP in this case, although there is some PERL backend). Right now, in order to keep from bogging a particular server down with latent or slow connections, I have had to impliment a load balanced configuration with multiple servers in a case where one or two could easily suffice if I could trust a multi-threaded model. We have actually discussed this a couple other times on the list. There is (at the very least) a percieved slowness in migration to the 2.0 apache setup and indeed slower to multi-threaded MPMs, even in situations that would absolutely benefit from them. Why? PHP is one of the biggest reasons I have been hearing from colleagues. Has any kind of extensive testing been done against a multi-threaded implementation of a PHP-based testing sled to find out what (if anything) is breaking in a threaded environment? Do we have hard data to offer the PHP community to encourage additional specific work and move towards thread-safe certification of some of the underlying PHP core and modules? - Wayne S. Frazee Any sufficiently developed bug is indistinguishable from a feature.
Re: OT: performance FUD
Paul A Houle wrote: Jess Holle wrote: So if one uses worker and few processes (i.e. lots of threads per), then Solaris should be fine? That's what people think, but I'd like to see some numbers. I've never put a worker Apache into production because most of our systems depend on PHP or something else which I wouldn't trust 100% in a threaded configuration. That's understandable if you're in that boat. We bundle and support our own Apache builds with our products. Our only dynamic content comes from mod_jk (and thus will come from the proxy AJP module in 2.2), so threading is all well and good. Given that most of our content is dynamic and thus via AJP, Apache performance is never really the issue -- if anything above the application code itself is ever an issue it is the extra hop involved with AJP, but there are clear load-balancing, security, etc, benefits from this architecture. Customers seem to consistently assume that using Apache is giving (substantively) lower overall performance than they'd get with something else, though -- chalk that up to good marketing by Microsoft, Sun, et al. As for the big file issue you note, that would only seem to be a big issue when coupled with slow connections -- which are getting rarer these days -- and much more of an issue with prefork than worker. -- Jess Holle
Re: OT: performance FUD
On 11/30/05, Colm MacCarthaigh [EMAIL PROTECTED] wrote: On Wed, Nov 30, 2005 at 11:01:55AM -0500, Paul A Houle wrote: So if one uses worker and few processes (i.e. lots of threads per), then Solaris should be fine? That's what people think, but I'd like to see some numbers. I've never put a worker Apache into production because most of our systems depend on PHP or something else which I wouldn't trust 100% in a threaded configuration. Is there anything we can do in 2.4/3.0 that will help gain that trust? I recently started a thread with (IMO) a potential solution: Any 'official' Apache FastCGI-like alternative planned? Even if PHP is thread-safe it's still not safe. One (bad) script can kill the entire server. The solution would be to run PHP (and other 'request processors') in a separate process.
Re: OT: performance FUD
On Wednesday 30 November 2005 16:30, Paul A Houle wrote: PHP's market position is as a product that any idiot can download and install, just following the instructions, and get a system with good reliability and performance -- a painful phase of shaking out threading bugs would endanger that perception. That looks a lot like Windows' market position. And I suspect it's no accident: both products have heaped on new 'goodies', all too often at the expense of other considerations. It's IMO also no accident that PHP is moving towards a Windows-like security track record. Which leads me to pose the question: can and should the PHP folks learn anything from how Microsoft are dealing with their tarnished image? And even, how closely should we @apache be watching, lest we ourselves stray from the straight-and-narrow and find ourselves at the wrong end of a bunch of real-life exploits, and/or get tarnished by fallout from elsewhere? -- Nick Kew
Re: Press release for httpd 2.2 (was Re: OT: performance FUD)
On Wed, Nov 30, 2005 at 11:25:06AM -0500, Joshua Slive wrote: Re Brian's question above, I think we want a statement emphasizing performance and scalability. It doesn't need to be extremely precise, spec-wise. Most people reading a press release wouldn't care. We just want to transmit the message that you use apache httpd in a very-high load situation and that it perfoms very well. By the way, are you using 2.0 or 2.1? Here's what we've come up with anyway; ftp.heanet.ie has been using Apache httpd 2.1/2.2 for over 6 months and has handled up to 27,000 concurrent downloads from a single webserver, while delivering terabytes of content per day. Large-file support, graceful-stop and mod_cache have improved our level of service dramatically. That hasn't been approved approved just yet, so please don't use it, but tomorrow when our PR person is back I'll hopefully get it, or something approximating it, approved. -- Colm MacCárthaighPublic Key: [EMAIL PROTECTED]
Re: Press release for httpd 2.2 (was Re: OT: performance FUD)
On Wed, Nov 30, 2005 at 05:13:07PM +, Colm MacCarthaigh wrote: On Wed, Nov 30, 2005 at 11:25:06AM -0500, Joshua Slive wrote: Re Brian's question above, I think we want a statement emphasizing performance and scalability. It doesn't need to be extremely precise, spec-wise. Most people reading a press release wouldn't care. We just want to transmit the message that you use apache httpd in a very-high load situation and that it perfoms very well. By the way, are you using 2.0 or 2.1? Here's what we've come up with anyway; ftp.heanet.ie has been using Apache httpd 2.1/2.2 for over 6 months and has handled up to 27,000 concurrent downloads from a single webserver, while delivering terabytes of content per day. Large-file support, graceful-stop and mod_cache have improved our level of service dramatically. That hasn't been approved approved just yet, so please don't use it, but tomorrow when our PR person is back I'll hopefully get it, or something approximating it, approved. I'll just remind everyone this is a public list and its archived too. =) If you wish to keep things private, we can use [EMAIL PROTECTED] and possibly the PMC list. Yes, there's a difference between including it in a PR and in an informal email, but you never know with press types. -- justin
Re: Press release for httpd 2.2 (was Re: OT: performance FUD)
Let my try to contribute my $.02: Usually, a PR item which tells about company X delivered product Y to customer Z, is signed by both - X and Z. Since most of the organizations don't exist for charity but for business, Z must earn something out of this PR. Sometimes, the fact that Z agreed to be an early adopter and even to put its name on the PR, gives Z a small discount (irrelevant in our case). But the usual case is that only by using the product Y, customer Z succeeded to execute its extraordinary service, and here come some impressive numbers which could not be achieved without Y, and so on. In other words, while the interest of X is clear, the interest of Z is more tricky, and usually Z uses this opportunity to tell the world how great he is. We owe Brian something, not only to pay him for putting the name of CNN here, but also to help him pass his legal staff. So this is, more or less, how I see the PR (not the announcement about 2.2, but specifically the CNN case) (and excuse me for my bad English...): -- Just an example, only to demonstrate what I mean -- After gaining more than 70% of the market (according to market research companies such as NetCraft and Security-Space), the future of the Apache web server looks brigher than ever, and it seems that it has no competitors anymore. But there is still one huge competitor that even Apache can't beat: Previous versions of Apache. It just works, 7 days a week, 24 hours a days, serving millions of requests, without losing even one says John Doe, a webmaster with the moc.com, which is ranked #0 among the hosting providers: After all, even Steve Ballmer said 'Apache is simply better', so why should I upgrade?! If it ain't broken, don't touch it!. But now, with Apache 2.2 coming soon, more and more people argue: We succeeded to hire the best reporters and journalists, but in order to keep our status as the most popular news site, it is not enough to create the best content, but we must find creative ways to deliver the enormous load required by our on-line readers says Brian Akins of CNN. Apache 2.2 allowed us the break even our own records, and reach an amazing number of 77 billion hits, although we started to use it only several days ago. I know no other way to deliver one billion pages per day summarizing Akins. This is the message that the Apache Software Foundation tries to make these days: Apache 1.3 is still the power behind most of the leading websites, but the new release combines the advantages of the old one with new fabulous features and abilities says XXX, a member of the Apache Software Foundation, and as an Open-Source product, it's free, so why not upgrade?!. What are these new features and abilities? XXX tries to shorten his answer, but the new release is so revolutionary, so the list looks infinite: (and here comes an infinite list of the goodies of Apache 2.2). The Apache web server is available for download from httpd.apache.org, free of charge. --- This was only an example, but I hope you got the idea. The only problem with this direction is that some of the newspapers and TVs which should publish it, are direct competitors of CNN, and may prefer to edit it (bad) or even to ignore this PR (bad too). -- Eli Marmor [EMAIL PROTECTED] Netmask (El-Mar) Internet Technologies Ltd. __ Tel.: +972-9-766-1020 8 Yad-Harutzim St. Fax.: +972-9-766-1314 P.O.B. 7004 Mobile: +972-50-5237338 Kfar-Saba 44641, Israel
Re: OT: performance FUD
Nick Kew wrote: That looks a lot like Windows' market position. And I suspect it's no accident: both products have heaped on new 'goodies', all too often at the expense of other considerations. It's IMO also no accident that PHP is moving towards a Windows-like security track record. You'll find skeletons if you go looking in CPAN. Market share is a lot of the reason why people target malware at Windows. If you wrote an email virus for the mac, one mac would infect the other mac and that would be the end of your fun. The real trouble with PHP is that it's sparked a revolution in web server software: code reuse. Before PHP, you couldn't find affordable web hosting for dynamic sites: cgi-bin was so expensive and problematic that mass hosting facilities couldn't afford to host it. Mod_perl would be out of the question. If you wanted to start a weblog or a wiki four years ago, you couldn't find reliable software that would hold up in the real world unless you were willing to put a lot of work in it. Today you can download Drupal, Wordpress or any of a large number of packages. So now there are tens of thousands of site running the same software with predictable URLs that people can mess around with and find bugs in the underlying software. If there were any Perl or Java apps of the same popularity, we'd be seeing the same thing. The difference is you can get a shared web hosting account for $10 / month if you want to run a Wordpress site on PHP, but you really want a dedicated server, more like $200 /month if you want to run mod_perl or Java. If you wanted to match the functionality of PHP, in mod_perl or Java, you'd have to install twenty or so framework modules -- everybody is going to pick a different set of modules, so attackers aren't going to have a consistent profile to hit, but on the other hand, this inconsistency makes it harder to incorporate other people's code into your site.
Re: OT: performance FUD
On Wednesday 30 November 2005 18:27, Paul A Houle wrote: Nick Kew wrote: That looks a lot like Windows' market position. And I suspect it's no accident: both products have heaped on new 'goodies', all too often at the expense of other considerations. It's IMO also no accident that PHP is moving towards a Windows-like security track record. You'll find skeletons if you go looking in CPAN. Market share is a lot of the reason why people target malware at Windows. That's offtopic and largely untrue. But a counterexample is ontopic: if market share were really the determinant, what webserver would Nimda, Code Red, Code Blue et al have hit? The truth is that Mac, Linux and others have *ample* market share. Even a real minority-system like RiscOS has been targeted. The real trouble with PHP is that it's sparked a revolution in web server software: code reuse. Like CGI.pm, libwww-perl, DBI/DBD et al never did? Erm Before PHP, you couldn't find affordable web hosting for dynamic sites: cgi-bin was so expensive and problematic that mass hosting facilities couldn't afford to host it. Mod_perl would be out of the question. I had no trouble finding cheap CGI hosting before moving to my own server. Actually that's not entirely true: my first host was not competently run. But moving to pair.net in IIRC May '96 got me something that worked well. If there were any Perl or Java apps of the same popularity, we'd be seeing the same thing. There are cultural reasons that are more important. If you post a Perl script that omits perl's taint checking, you'll predictably get flamed for it. So the newbie programmer has to figure out whats going on, and take in some basic principles of security in the process. Of course you can't guarantee that'll work, but it makes a better environment for safety- awareness than PHP. And it's not as if Perl is, in general, something I'd hold up as a role model for good practice, either:-) -- Nick Kew
Re: Press release for httpd 2.2 (was Re: OT: performance FUD)
On Wed, Nov 30, 2005 at 09:19:25AM -0800, Justin Erenkrantz wrote: I'll just remind everyone this is a public list and its archived too. =) If you wish to keep things private, we can use [EMAIL PROTECTED] and possibly the PMC list. Yes, there's a difference between including it in a PR and in an informal email, but you never know with press types. -- justin Thanks, and just to make the prc aware of one more thing; our release of 2.2.0 will be done ten years to the day (well in as much as a day can be put on it) since Apache 1.0.0. The first ever Apache httpd GA. There may well be some PR in that too :) -- Colm MacCárthaighPublic Key: [EMAIL PROTECTED]
Re: Press release for httpd 2.2 (was Re: OT: performance FUD)
Colm MacCarthaigh wrote: Thanks, and just to make the prc aware of one more thing; our release of 2.2.0 will be done ten years to the day (well in as much as a day can be put on it) since Apache 1.0.0. The first ever Apache httpd GA. There may well be some PR in that too :) I feel old... :) -- === Jim Jagielski [|] [EMAIL PROTECTED] [|] http://www.jaguNET.com/ If you can dodge a wrench, you can dodge a ball.
Re: Press release for httpd 2.2 (was Re: OT: performance FUD)
On Wed, Nov 30, 2005 at 11:52:38AM -0800, Justin Erenkrantz wrote: --On November 30, 2005 7:39:08 PM + Colm MacCarthaigh [EMAIL PROTECTED] wrote: Thanks, and just to make the prc aware of one more thing; our release of 2.2.0 will be done ten years to the day (well in as much as a day can be put on it) since Apache 1.0.0. The first ever Apache httpd GA. There may well be some PR in that too :) http://mail-archives.apache.org/mod_mbox/httpd-dev/199512.mbox/[EMAIL PROTECTED] Figures that it'd take us 10 years to go from 1.0-2.2. ;-) -- justin It's a great display of our commitment to stability, long-term strategic planning and thinking, enterprise-grade development and trustworthy convervatism. Right? ;-) -- Colm MacCárthaighPublic Key: [EMAIL PROTECTED]
Re: Press release for httpd 2.2 (was Re: OT: performance FUD)
On Nov 30, 2005, at 11:56 AM, Colm MacCarthaigh wrote: On Wed, Nov 30, 2005 at 11:52:38AM -0800, Justin Erenkrantz wrote: --On November 30, 2005 7:39:08 PM + Colm MacCarthaigh [EMAIL PROTECTED] wrote: Thanks, and just to make the prc aware of one more thing; our release of 2.2.0 will be done ten years to the day (well in as much as a day can be put on it) since Apache 1.0.0. The first ever Apache httpd GA. There may well be some PR in that too :) http://mail-archives.apache.org/mod_mbox/httpd-dev/199512.mbox/% [EMAIL PROTECTED] Figures that it'd take us 10 years to go from 1.0-2.2. ;-) -- justin It's a great display of our commitment to stability, long-term strategic planning and thinking, enterprise-grade development and trustworthy convervatism. Right? ;-) Yes and we should milk that for all it's worth. S. -- [EMAIL PROTECTED] http://www.temme.net/sander/ PGP FP: 51B4 8727 466A 0BC3 69F4 B7B8 B2BE BC40 1529 24AF
Re: Press release for httpd 2.2 (was Re: OT: performance FUD)
On Wednesday 30 November 2005 20:09, Sander Temme wrote: It's a great display of our commitment to stability, long-term strategic planning and thinking, enterprise-grade development and trustworthy convervatism. Right? ;-) Yes and we should milk that for all it's worth. There was talk of an Apachecon-Stuttgart release date. The extra six months demonstrates our committment to quality and stability ... ... or could look like a Microsoft release schedule. -- Nick Kew
Re: OT: performance FUD
Justin Erenkrantz wrote: --On November 29, 2005 3:40:11 PM -0500 Paul A Houle [EMAIL PROTECTED] wrote: * prefork and worker seem to be about equally fast on linux? * MacOS X? fork() is *painfully* slow on the darwin kernel, I haven't tested but can't imagine that threading isn't a huge win here.
Re: OT: performance FUD
Brian Akins wrote: Colm MacCarthaigh wrote: I've never put a worker Apache into production because most of our systems depend on PHP or something else which I wouldn't trust 100% in a threaded configuration. Is there anything we can do in 2.4/3.0 that will help gain that trust? PHP, or it's extensions or whatever they call them, are not thread safe. So until that's fixed, nothing we can do. Probably the same with other stuff. I don't suppose we should respond to the fud that PHP is 20th century technology that isn't compatible with 21st century high-availablity operating systems? Seems like it would be in poor taste :) Someone once suggested a list of the 'minimum library versions' for thread safety and reentrant support (note the 2nd may be required as the async httpd server evolves.) Essentially without that knowledge, although PHP is mostly thread safe, the extentions are unknowns. Bill
Re: Press release for httpd 2.2 (was Re: OT: performance FUD)
Justin Erenkrantz wrote: --On November 30, 2005 7:39:08 PM + Colm MacCarthaigh [EMAIL PROTECTED] wrote: Figures that it'd take us 10 years to go from 1.0-2.2. ;-) -- justin Good point, I propose we call this coming release 3.0 :)
Re: Press release for httpd 2.2 (was Re: OT: performance FUD)
William A. Rowe, Jr. wrote: Good point, I propose we call this coming release 3.0 :) No It should be Apache HTTP X (ten). All the other cool things are at 10 (or close). Mac OS, Solaris, Suse, Redhat, Mandrake -- Brian Akins Lead Systems Engineer CNN Internet Technologies
Re: OT: performance FUD
--On November 29, 2005 2:50:19 PM -0500 Brian Akins [EMAIL PROTECTED] wrote: This is probably way off topic for this list. I was searching for something related to php this morning (I know, I know... But some people here need php) and the majority of the google hits where FUD sites. Most of them generally say Apache is bloated and slow, you should use X. I know we have several people on this list who run Apache on very high traffic sites. While we cannot answer every single piece of FUD out there, do we need a general page to answer some of them. Maybe testimonials or something. I know, with my config, I can easily saturate multiple gig interfaces and have a rather full feature installation. Thoughts? First off, this is absolutely on-topic for this list. =) Most high-traffic sites keep their details under wraps. If you are willing to have some information made public (i.e. how much traffic you do, how many servers, etc.), I'm sure we could post a page on our website towards that end. I've presented some talks about how apache.org runs httpd with one small httpd instance. It'd be nice to complement that information with other sites who have far more complicated setups. -- justin
Re: OT: performance FUD
Justin Erenkrantz wrote: Most high-traffic sites keep their details under wraps. If you are willing to have some information made public (i.e. how much traffic you do, how many servers, etc.), I'm sure we could post a page on our website towards that end. Would this be a worthwhile ApacheCon topic? (I'm already angling to get a speakers spot for next year.) We do some interesting things here :) For the time being, for a web page, I could discuss some performance metrics we use. The amount of traffic we get is basically public knowledge (we have press releases from time to time). I'm sure I could talk/write in generalities, but not we use X servers. But I can certainly report some general benchmark numbers. Does that make any sense? -- Brian Akins Lead Systems Engineer CNN Internet Technologies
Re: OT: performance FUD
Justin Erenkrantz wrote: It'd be nice to complement that information with other sites who have far more complicated setups. -- justin This could also be part of a press release announcing 2.2. Just between Brian and Colm we could have a couple impressive-sounding quotes. Joshua.
Re: OT: performance FUD
--On November 29, 2005 3:16:43 PM -0500 Joshua Slive [EMAIL PROTECTED] wrote: This could also be part of a press release announcing 2.2. Just between Brian and Colm we could have a couple impressive-sounding quotes. Agreed. If they're up for being quoted, they'd be great. -- justin
Re: OT: performance FUD
Joshua Slive wrote: This could also be part of a press release announcing 2.2. Just between Brian and Colm we could have a couple impressive-sounding quotes. I know that a press release is out of the question for my company. We do not endorse or disparage any product. We could be referred to as one large company or something similar... It's not that hard to figure out what we run.. -- Brian Akins Lead Systems Engineer CNN Internet Technologies
Re: OT: performance FUD
On Tue, Nov 29, 2005 at 03:16:43PM -0500, Joshua Slive wrote: This could also be part of a press release announcing 2.2. Just between Brian and Colm we could have a couple impressive-sounding quotes. +1 great idea! Making it known that the 2.2GA has been stress tested at high profile sites for quite a while is likely to help along 2.2 adoption. vh Mads Toftum -- `Darn it, who spiked my coffee with water?!' - lwall
Re: OT: performance FUD
--On November 29, 2005 3:15:05 PM -0500 Brian Akins [EMAIL PROTECTED] wrote: Would this be a worthwhile ApacheCon topic? (I'm already angling to get a speakers spot for next year.) We do some interesting things here :) For Absolutely. the time being, for a web page, I could discuss some performance metrics we use. The amount of traffic we get is basically public knowledge (we have press releases from time to time). I'm sure I could talk/write in generalities, but not we use X servers. But I can certainly report some general benchmark numbers. Does that make any sense? Sure. FWIW, most people only care about generalities not any specifics. We don't need to have specifics. Stuff like Just under 1 billion requests here is spot-on. =) -- justin
Re: OT: performance FUD
Justin Erenkrantz wrote: Sure. FWIW, most people only care about generalities not any specifics. We don't need to have specifics. Stuff like Just under 1 billion requests here is spot-on. =) -- justin Cool. I have asked my SVP what is acceptable for me to say. But it doesn't take a rocket scientist to figure it out: http://toolbar.netcraft.com/site_report?url=http://www.cnn.com plus http://www.lostremote.com/archives/002933.html -- Brian Akins Lead Systems Engineer CNN Internet Technologies
Re: OT: performance FUD
Brian Akins wrote: This is probably way off topic for this list. I was searching for something related to php this morning (I know, I know... But some people here need php) and the majority of the google hits where FUD sites. Most of them generally say Apache is bloated and slow, you should use X. I know we have several people on this list who run Apache on very high traffic sites. While we cannot answer every single piece of FUD out there, do we need a general page to answer some of them. Maybe testimonials or something. I know, with my config, I can easily saturate multiple gig interfaces and have a rather full feature installation. Apache isn't the fastest web server -- at least without mod_event. I've seen data corruption with all of the free single-process web servers, although I'd assume that products like Zeus do better. Looking at Alexa, the logs from a few sites I run, and benchmarking I've done, there are probably only a few thousand web sites in the world that push the limits of a single Apache web server. Perhaps 100x as many PHB's ~might~ pick a web server because of numbers in a glossy ad. The real competition is with IIS, and people don't choose Apache or IIS based on performance numbers -- they choose it because they are familiar with Unix or familiar with Windows. Other web servers are at the 1% market share level: http://news.netcraft.com/archives/2005/11/07/november_2005_web_server_survey.html Don't make it a fudbusting site, make it a apache performance tuning site. There are all of these statements in the apache docs that * .htaccess is slow * ExtendedStatus on reduces performance We did a round of performance testing on a server that we commissioned last year and took measurements of these things, and found that we'd need to put 1000 rewriting rules to harm performance noticably, that the overhead of ExtendedStatus On is negligible for a site that gets 500 hits/sec, etc. I might see if I can find my report about this on this and put it online -- there some things that I know, and even more that I don't... * prefork and worker seem to be about equally fast on linux? * is the case on Solaris? * MacOS X? * Solaris 9 is embarassingly slow running Apache compared to Linux -- is the same the case with Solaris 10?
Re: OT: performance FUD
--On November 29, 2005 3:40:11 PM -0500 Paul A Houle [EMAIL PROTECTED] wrote: * prefork and worker seem to be about equally fast on linux? Yup - this is because forking and threading are equivalent (by and large) on Linux. * is the case on Solaris? No. The threading gains of worker are seen with Solaris and, I believe, AIX. * MacOS X? I haven't done a lot of high-end sites on Mac OS X. * Solaris 9 is embarassingly slow running Apache compared to Linux -- is the same the case with Solaris 10? If it's on equivalent hardware (i.e. Linux/Intel vs. Solaris/Intel on the same box), I doubt there will be an extreme performance gap. In fact, I've often seen Solaris outperform Linux on certain types of loads. In my experience, a lot of Linux network card drivers are sub-standard; if it's supported by Solaris, there's a fair chance the driver takes full advantage of the hardware. (Netgear GigE drivers on Linux are abysmal.) -- justin
Re: OT: performance FUD
Paul A Houle wrote: Don't make it a fudbusting site, make it a apache performance tuning site. There are all of these statements in the apache docs that * .htaccess is slow * ExtendedStatus on reduces performance We did a round of performance testing on a server that we commissioned last year and took measurements of these things, and found that we'd need to put 1000 rewriting rules to harm performance noticably, that the overhead of ExtendedStatus On is negligible for a site that gets 500 hits/sec, etc. I might see if I can find my report about this on this and put it online -- there some things that I know, and even more that I don't... * prefork and worker seem to be about equally fast on linux? * is the case on Solaris? * MacOS X? * Solaris 9 is embarassingly slow running Apache compared to Linux -- is the same the case with Solaris 10? Suggestions to improve http://httpd.apache.org/docs/2.1/misc/perf-tuning.html are very welcome. Suggestions backed by data are even better. One issue is that this page was written for (and, in fact, by) the Dean Gaudet-type performance freak who was looking to squeeze every last ounce of performance when serving static pages. All you need to do is add one CGI script or php app to your site and everything on that page after the hardware section gets lost in the noise. So when people mail [EMAIL PROTECTED] asking how to fix performance problems, the answer is almost always fix your database or rewrite your web app and not change your apache configuration or get a faster web server. That is why just communicating the fact that apache is fast enough in almost all cases is important. Joshua.
Re: OT: performance FUD
Joshua Slive wrote: Suggestions to improve http://httpd.apache.org/docs/2.1/misc/perf-tuning.html are very welcome. Suggestions backed by data are even better. Basically there's nothing quantitative there. There's a lot of talk about some operating systems and not a lot of talk about specifics. One issue is that this page was written for (and, in fact, by) the Dean Gaudet-type performance freak who was looking to squeeze every last ounce of performance when serving static pages. All you need to do is add one CGI script or php app to your site and everything on that page after the hardware section gets lost in the noise. So when people mail [EMAIL PROTECTED] asking how to fix performance problems, the answer is almost always fix your database or rewrite your web app and not change your apache configuration or get a faster web server. For me, that's the reason why quantitative information is so important. I did extensive performance testing on the new server we commissioned precisely because of the situation you describe: we had people saying rewriting is slow, extendedstatus on is slow -- people were making decisions based on qualitative statements about performance, not qualitative performance. After doing those tests, I learned that I had nothing to fear if I wanted to put in 500 rewriting rules, but that 50,000 is too much.
Re: OT: performance FUD
Justin Erenkrantz wrote: If it's on equivalent hardware (i.e. Linux/Intel vs. Solaris/Intel on the same box), I doubt there will be an extreme performance gap. In fact, I've often seen Solaris outperform Linux on certain types of loads. In my experience, a lot of Linux network card drivers are sub-standard; if it's supported by Solaris, there's a fair chance the driver takes full advantage of the hardware. (Netgear GigE drivers on Linux are abysmal.) -- justin I think the issue with Apache/Solaris is that process switches take a long time on Solaris. I've got a computer that I need to rehabilitate for the family this Christmas. I think I'll put Solaris 10 on it and do some web server benching before I put Ubuntu on it.
Re: OT: performance FUD
On Tuesday 29 November 2005 20:49, Justin Erenkrantz wrote: --On November 29, 2005 3:40:11 PM -0500 Paul A Houle [EMAIL PROTECTED] wrote: * prefork and worker seem to be about equally fast on linux? Yup - this is because forking and threading are equivalent (by and large) on Linux. That's the conventional wisdom for static stuff. If DBD catches on it'll change. In my experience, a lot of Linux network card drivers are sub-standard; if it's supported by Solaris, there's a fair chance the driver takes full advantage of the hardware. (Netgear GigE drivers on Linux are abysmal.) -- justin Is that why Linux server (as opposed to desktop) hardware tends to standardise on a few well-supported brand such as 3Com and Intel? I have 3Com on the server and something much cheaper on the desktop:-) And I discovered the hard way that SCSI discs are still a Good Idea if they're doing something nontrivial in the server. -- Nick Kew
Re: OT: performance FUD
On Tue, Nov 29, 2005 at 03:18:57PM -0500, Brian Akins wrote: I know that a press release is out of the question for my company. We do not endorse or disparage any product. That's understandable, for a news organisation. For our part, HEAnet has no problem being quoted, but maybe something like; Apache httpd 2.2 has been developed with community involvement and has been deployed and tested on some of the most active websites on the internet, and has handled billions of requests in production environments. We could be referred to as one large company or something similar... It's not that hard to figure out what we run.. It's public knowledge that CNN.com runs Apache 2.2, would it cause you a lot of trouble for that to be referenced? -- Colm MacCárthaighPublic Key: [EMAIL PROTECTED]
Re: OT: performance FUD
Paul A Houle wrote: Justin Erenkrantz wrote: If it's on equivalent hardware (i.e. Linux/Intel vs. Solaris/Intel on the same box), I doubt there will be an extreme performance gap. In fact, I've often seen Solaris outperform Linux on certain types of loads. In my experience, a lot of Linux network card drivers are sub-standard; if it's supported by Solaris, there's a fair chance the driver takes full advantage of the hardware. (Netgear GigE drivers on Linux are abysmal.) -- justin I think the issue with Apache/Solaris is that process switches take a long time on Solaris. So if one uses worker and few processes (i.e. lots of threads per), then Solaris should be fine? -- Jess Holle
Re: OT: performance FUD
On Tuesday 29 November 2005 21:28, Colm MacCarthaigh wrote: On Tue, Nov 29, 2005 at 03:18:57PM -0500, Brian Akins wrote: I know that a press release is out of the question for my company. We do not endorse or disparage any product. That's understandable, for a news organisation. For our part, HEAnet has no problem being quoted, but maybe something like; Hmmm, how about an early adopters page? We could *imply* the organisations by featuring mugshots and brief profiles of both Brian and Colm as having successfully beta-tested 2.1.x in very-high-volume production environments. Just a thought :-) -- Nick Kew