Mike Brittain <[EMAIL PROTECTED]> writes: > > I'm trying to review server logs to determine the number of completed > downloads of an installation program available from a server. I have > already reduced the reported requests to just the name of the file I'm > reporting on: > > FILEEXCLUDE * > FILEINCLUDE *my_installer.exe > > When I review the file size report, I'm seeing that the file is reported > as a variety of sizes. And reviewing the entries in the original log > file supports this. Only a subset of all requests to that file have the > full file size in bytes associated with the request line. Many of the > requests have a delivered file size reported as being smaller than the > original file, which I assume indicates an aborted download. Sample > file size report (actual file size is about 1.8 MB): > > size reqs %bytes > 1B- 10B 0 > 11B- 100B 0 > 101B- 1kB 0 > 1kB- 10kB 0 > 10kB-100kB 14 1.85% > 100kB- 1MB 9 4.62% > 1MB- 10MB 24 93.53% > > Is there any way to report /completed/ downloads, as opposed to just all > requests for a file, whether the entire file was delivered or not? > > Thanks, > Mike >
I am not an expert at this, but I have a similar problem. Looking at my acces log I noticed that in some cases I would get at status code 200 followed by the full file size. In other cases, I would get a status code 200 immediately followed by several status codes 206 from the same browser, presumeably requesting the download in smaller (partial) parts. I have resorted the using STATUSEXCLUDE 206 to stop counting the multiple partial requests for the same file. It appears I get at least one 200, then maybe several 206s. So counting the 200s seem to get me a more accurate count. Rick +------------------------------------------------------------------------ | TO UNSUBSCRIBE from this list: | http://lists.meer.net/mailman/listinfo/analog-help | | Usenet version: news://news.gmane.org/gmane.comp.web.analog.general | List archives: http://www.analog.cx/docs/mailing.html#listarchives +------------------------------------------------------------------------
