#php
08 October 2007
Total 41 pages. You are browsing page 1/41.
First :: Prev :: [1] [2] [3] [4] [5] [...] :: Next :: Last
--- Log opened Mon Oct 08 00:00:59 2007
00:01 <****> "onclick=system("calc.exe");" Billium that looks like javascript too me?
00:03 <****> do you want to open a program on your server, or on the client?
00:04 <****> How reliable are bogomips when it comes to getting an approximate gauge of CPU power?
00:04 <****> SpNg: Why not? IPs are IPs. Assuming you have the appropriate routes, you'd be fine.
00:05 <****> xtor: a good benchmark will do the job
00:05 <****> it depends mostly on the apps you plan to run
00:05 <****> are there any sites that have the real benchmarks or MIPs..for running active content PHP on an apache box?
00:06 <****> xtor: your benchmarks are only in real live scenarious
00:06 <****> or stress testing your site
00:06 <****> xtor: look at ab [apache bench]
00:07 <****> Assid but I need to know how much better a certain CPU will be if I upgrade
00:07 <****> mattmcc: well one IP is the one our nameserver resolves, the other would be the local IPs of our servers...
00:07 <****> xtor: do you use a sql database for your site?
00:07 <****> xtor: it's highly likely that your cpu will not be any sort of bottleneck
00:07 <****> no SQL used just PHP
00:08 <****> I dont even use squid yet
00:08 <****> xtor: then honestly.. i dont think your gonna max out a quad
00:08 <****> not unelss you got some crazy codes in there
00:08 <****> which read binary files in some way
00:08 <****> Assid I got an E6600 server in mind...with 4MB L2 cache...is that good?
00:08 <****> crappy applications are more likely to be the cause :) getting a faster server only makes crappy applications go a little bit faster, it doesn't stop them from being crap
00:09 <****> how to increase the number of php connections allowed by the server?
00:09 <****> yeah so you are saying I should rewrite that PHP script myself I guess
00:09 <****> xtor: yeah..
00:09 <****> im actually on the Q6600 which is the E6600 bigger brother
00:09 <****> I'm saying you should find out what is likely to be the bottleneck before pre-optimizing and wildly guessing
00:10 <****> HarryR: thats sooooooooooooo true
00:10 <****> :)
00:10 <****> HarryR: i have a client that used to refuse the idea his code sucks
00:10 <****> Assid does your load average ever go over 1 during peak hours?
00:10 <****> same as, many times
00:10 <****> xtor: mine peaks at maybe 7 or 8
00:10 <****> finally i told him TRY it at his local machine.. he finally saw his asp.net script didnt close the sql connections..
00:10 <****> 3 or 4 is moderate load...
00:11 <****> xtor: oh this is my desktop.. but i have a xeon 3060 .. and loads are like 2-3 average.. and thats with TONS of email being scanned with spamassassin and clamd
00:11 <****> hrmm thats disturbing...do you run a lot of MYSQL
00:11 <****> Oh I thought it was a dedicated server
00:12 <****> xtor: every email that goes through the system is recording in mysql.. my ftp daemon has a mysql backend
00:12 <****> my load avg goes up to the 20s at times :P
00:12 <****> usually when something is wrong.
00:12 <****> :)
00:12 <****> JonathanD: yeah that would be a sign of something going wrong
00:12 <****> lol
00:12 <****> xtor: nah dedicated server is only a xeon 3060
00:12 <****> Assid: it's still functional at 20
00:12 <****> I've seen an average of ~20-25 on a big sun box and it was well within it's limits
00:13 <****> when it hits about 60-70, it stops responding
00:13 <****> I am attempting to use " while (!feof($InputFileExists)) " to parse a file and act on it's contents, which, in this case, is to register user accounts based on an input file of names etc. It all works fine, but just at the end it adds an empty user account? I'm not sure why though. I suspect it may have something to do with fclose but I have that at the end after my while loop. http://pastebin.com/d1d439ff4 .
00:13 <****> JonathanD: yes.. ive recovered a server back from loads of 137 and above
00:13 <****> it's all relative to your machine
00:13 <****> Assid: if I was more patient, maybe :P
00:13 <****> JonathanD: not really.. it becomes suppppppppper slow.. but you can still get things back.. you just gotta wait a whole lot longer
00:13 <****> init 1 ;)
00:14 <****> it's at 7 right now.
00:14 <****> on the weekend, at that
00:14 * JonathanD pats his wonderful little box
00:14 <****> what do you have going through that ?!
00:14 <****> you oversold your hosting box?
00:14 <****> me?
00:14 <****> vmware
00:14 <****> ok
00:14 <****> ah
00:14 <****> that would explain it
00:14 <****> does L2 cache help on a dedi much? Like going from 2MB L2 to 4MB L2?
00:14 <****> xtor: for things like php/apache ye
00:14 <****> xtor: it depends on the apps your running
00:15 <****> generally more applications and concurrent threads can do with more L2 cache
00:15 <****> higher cache generally means quicker io
00:15 <****> assid: l2 cache has little to do with io...
00:15 <****> io being CPU to RAM io though
00:15 <****> cuz it just caches RAM I gather
00:15 <****> HarryR: the number of aplications is correlated to the io passing through the system
00:15 <****> i dont mean the hard drive per se
00:15 <****> just how much code it can cache on-die
00:16 <****> i mean the system/cpu itself
00:16 <****> ya
00:16 <****> Assid: it's a whole environment :) 3 apache/php servers, 2 mysql, DNS, a few other things.
00:16 <****> JonathanD: p4 ?
00:16 <****> Assid: dual xeon.
00:16 <****> DS4700 for storage.
00:16 <****> maybe time to upgrade?
00:16 <****> or split them to smaller boxes?
00:16 <****> this is test/dev, mostly.
00:16 <****> i've seen standing loads of around 0.5 to 1 on an empty vmware machine
00:17 <****> so 7 is nothing... average load :)
00:17 <****> my desktop has no loads.. so.. i just fold - boinc
00:17 * HarryR wishes PHP was more effecient
00:17 <****> sigh :( http://paste2.org/p/8059 trying to use the system() function
00:17 <****> and I need a native kcackegrind for cocoa :(
00:17 <****> HarryR: its not THAT bad
00:18 <****> assid: really, it is compared to Java :(
00:18 <****> really?
00:18 <****> 5-10x slower in my development environment
00:19 <****> and 5x faster running PHP on Caucho Quercus
00:19 <****> i don't want to threat in other page
00:19 <****> in php, whats the difference in using " " and ' '?
00:19 <****> HarryR: APC?
00:19 <****> HarryR: ?
00:19 <****> mattmcc: i've been using eAccelerator, but it's still slower than Java
Total 41 pages. You are browsing page 1/41.
First :: Prev :: [1] [2] [3] [4] [5] [...] :: Next :: Last