SARG 2.3.2 on Windows Info report Top Sites

sarg
2012-05-17
2012-12-19
  • roberto cesani

    roberto cesani - 2012-05-17

    hello,
    i need help for report "Top Sites".
    in the report appear some IP unknown eg. 92.43.21.136:443 or 213.199.179.165:443 and some other.
    how can I do to not visualize these IP on the report.
    I try with tag exclude_hosts eg. 92.43.21.136 but IP is still on report.

    thanks
    roberto

     
  • Frederic Marchal

    As you found out, you have to use exclude_hosts with a file name.

    The file you specify is a simple text file to be created by yourself. It contains the hosts you want to exclude from the reports (one per line). Each line may contain a IPv4 address; a IPv4 subnet such as 10.0.0.0/8; a url or url starting with a wildcard (the wildcard is * and must be the first character of the url). IPv6 addresses are not accepted by sarg 2.3.2 but that's not your question…

    Back to your problem. Run sarg with the two additional command line options -x -z. Look at the output for a line such as

    SARG: Loading exclude host file from: /etc/sarg/sarg.hosts
    

    Make sure sarg is reading the file you expect.

    Frederic

     
  • roberto cesani

    roberto cesani - 2012-05-21

    Frederic, thanks for your time.
    I tried your instructions, this is the output
    the txt file in the file sarg.conf is exactly exclude_hosts.txt .
    in the file exclude_hosts.txt there are only 3 rows
    192.0.0.0
    213,146,189,206
    *. ebay.it
    in the file squid access.log there are occurrences with www.ebay.it and 213,146,189,206.
    into the report do not appear www.ebay.it  but appear 213,146,189,206.
    where is my mistake.

    regards.
    roberto

     
  • Frederic Marchal

    Are you sure www.ebay.it would have appeared in the report had it not been excluded by exclude_hosts.txt? I ask this because the excerpt you posted says "*. ebay.it" with a space between the first dot and "ebay.it". It may be a typo in your post but if it is indeed spelled like this in exclude_hosts.txt, it means it is not filtered out by the exclude_hosts.txt file and your test is not conclusive.

    In addition, www.ebay.it could occur so infrequently as to be ignored during the report generation. Can you check this by just removing that line from exclude_hosts.txt (you can prefix it with # to comment it out) and see if the url appears in the report?

    Then, if you are sure the exclude_hosts.txt is taken into account (because you see www.ebay.it in the report when it is not excluded and it is not in the report when it is excluded) then it may be a bug reported today in another post. If the IP address is followed by a port number (for instance 213.146.189.206:443), then the filtering out won't work no matter what you try.

    If it is the case, I can try to package the latest sarg for you to give it a try.

    One more guess: you wrote the IP address with commas instead of dots. I assume you wrote it in your post like this for convenience but the IP address must be written with dots in the exclude_hosts.txt.

    Frederic

     
  • roberto cesani

    roberto cesani - 2012-05-22

    Hello Frederic,
    sorry for my bad English.
    the site www.ebay.it is correctly excluded. The filter with exclude_hosts works.
    the problem is IP 213.146.189.206:443.
    I'm not able to exclude this IP.
    in files exclude_hosts there is a row with 213.146.189.206 but does not work.
    If you prepare a patch I'll be happy to try it.

    regards
    roberto

     
  • Frederic Marchal

    Thanks for your detective work Roberto. You are definitely seeing the bug reported yesterday.

    I uploaded a new build for windows to solve your problem. It is version 2.3.3-pre1. Well, to make things clear, it is a build of the source before I added a last minute feature (namely to run an executable to resolve the IP addresses into host names) but it won't change anything for you. You can download it, replace your current version (don't overwrite your configuration files with those in the zip file) and the IP addresses should be excluded from the report.

    You will be testing the latest development version before the first release candidate is even out. I'm interested in hearing about any bug or problem you encounter.

    Frederic

     
  • roberto cesani

    roberto cesani - 2012-05-22

    Hello Frederic,
    I tested the version 2.3.3 you have indicated.
    everything is working correctly.

    thanks
    roberto

     

Log in to post a comment.