On Friday, March 10, 2006 2:36 AM [EDT], analog-help at talent-aid.org <[EMAIL PROTECTED]> wrote:
>>> I have not specified any log format so it is whatever is the > default. >> >> If you're letting Analog recognize the logformat itself, it will >> ignore the method (GET, HEAD, OPTION, etc), so they won't be >> recognized in any ALIAS. > > You mean HEAD, OPTION No, I mean that analog ignores the method string in your logfile. That includes GET. It doesn't care whether the method is GET, POST, HEAD or OPTION, it will treat it s a successful request if the status code is 200, and it will treat it as a failed request if the status code is 40x. >>> I SORT of solved it with a >>> PAGEEXCLUDE / >>> >> You haven't really told us what it is that you have solved - what >> were you trying to achieve? Judging by your FILEALIAS examples, I >> guessed that you wanted to identify your own activities, but that >> doesn't make sense with a PAGEEXCLUDE. Note that PAGEEXCLUDE simply tells Analog to treat a specific document type as a "non-page" request. It's still counted as a request. (By default only .htm/.html requests are counted as pages). > We have a heart beat monitor that pings every so often. It does that > with a > GET http://www.myserver.com/ > > I assume that translates into the 111.222.333.444.555 GET / I see in > the log file (standard apache http log file) where 111.. and so on are > our own IP Then what you want is to HOSTEXCLUDE that IP address. All requests from that IP address will be ignored, but it's unlikely that there is any real traffic from that address. > But our homepage IS a page so I wanted not to page exclude it to begin > with. However we are not the only heartbeat monitors out there (we > have quite a few leechers out there). So I realise I have to > PAGEEXCLUDE the GET / or it will give a ridiculous page count. Ideally, you should HOSTEXCLUDE the relevant addresses. It should be possible to pick out the leechers by generating a Host Report, and sorting it by Requests, while displaying both the requests made and the pages requested. "Real" browsers always make many more requests than page-requests (images, CSS, etc), whereas spiders almost always have an equal number of requests and page requests Create a report with just these lines added to your analog.cfg: ALL OFF HOST ON HOSTCOLS RP > I am doing a host exclude > >> >>> >>> ... RealUsers: XXXXXXX requests >>> /scripts/css_ie.css xxxxx >>> /scripts/css_gecko.css xxxx >>> /scripts/css_ns_mac.css xxx >>> /scripts/css_gecko_mac.css xx >>> /scripts/css_ns.css x >>> >>> Is that possible? >> >> The default SORTORDER is by the number of requests. You can change >> the SORTORDER, but that will effect teverything else too. Probably >> the best way to achieve what you want to do in this specific >> instance would be to FILEALIAS the CSS files to a single file with >> the browser type as a parameter - something like >> >> FILEALIAS /scripts/css_*.css /scripts/css_browser.css?$1 > > Excellent idea > >> >> Or you could just look at the Browser Report :-). > > Not available - and that would be chock full of fake user agents > anyway. The leechers do not fake the useragents much on the css files > because they do not load the css > > Thanks a lot > > Michel > PS: You are very guruific, did you (co)write Analog or ReportMagic? I've been using Analog for years, and have been contributing to the mailing list since it started at the end of 1998. I've made suggestions for Analog, but I've never even read the source code, never mind writing it :-). > PPS: It is a huge pain going via the list (webmailer), if I am not > abusing you, may I email you directly? I would prefer that you use the list. Aengus +------------------------------------------------------------------------ | TO UNSUBSCRIBE from this list: | http://lists.meer.net/mailman/listinfo/analog-help | | Analog Documentation: http://analog.cx/docs/Readme.html | List archives: http://www.analog.cx/docs/mailing.html#listarchives | Usenet version: news://news.gmane.org/gmane.comp.web.analog.general +------------------------------------------------------------------------
