logging customizable?

This site uses cookies. By continuing to browse this site, you are agreeing to our Cookie Policy.

  • RE: logging customizable?

    Hello!
    You mean customizing log level for "palo.log", or did you mean something else? Currently there are only two generic log levels for palo.log - standard and detailed. Release version (the one you have) is the standard, less detailed version. Future Palo versions will probably support more options for customizing log data.
    Mit freundlichen Gruessen/ With kind Regards / Meilleures salutations

    Vladislav Malicevic
    VP Development & Support
    Jedox AG
  • RE: logging customizable?

    Hello,

    I'll describe my test setup for clarification:

    Source Code

    1. $ cd /opt/palo/server/data
    2. $ ls -l
    3. -rw-rw-r-- 1 palo palo 186 Feb 23 16:21 auth.xml
    4. -rw-rw-r-- 1 palo palo 291 Feb 23 16:20 config.xml
    5. drwxrwsr-x 2 palo palo 1024 Feb 27 11:39 test
    6. -rw-rw-r-- 1 palo palo 28436 Mar 7 10:35 palo.log
    7. $ cd test
    8. $ ls -l
    9. -rw-rw-r-- 1 palo palo 402 Mar 7 11:39 U0FMRVMyMDA1.cube.data
    10. -rw-rw-r-- 1 palo palo 202 Mar 7 11:37 U0FMRVMyMDA1.cube.log
    11. -rw-rw-r-- 1 palo palo 128 Mar 7 11:37 U0FMRVMyMDA1.cube.xml
    12. -rw-rw-r-- 1 palo palo 1066 Mar 7 11:37 U1RPUkU=.dimension.log
    13. -rw-rw-r-- 1 palo palo 504 Mar 7 11:39 U1RPUkU=.dimension.xml
    14. -rw-rw-r-- 1 palo palo 1059 Mar 7 11:37 UFJPREdST1VQ.dimension.log
    15. -rw-rw-r-- 1 palo palo 500 Mar 7 11:39 UFJPREdST1VQ.dimension.xml
    Display All
    My small example contains a database named "test" with two dimensions and one cube. As you can see, each of these consist of a .data file, a.log file, and a .xml file.

    I mean the .log files in the test directory. For example, adding one single value to the cube produces following log line in U0FMRVMyMDA1.cube.log:

    Source Code

    1. 1141036621.31 PALO SETDATA Augsburg Flutes double splash_disable 22.00000000000000000000000000000000
    That's 100 bytes.

    And now think of a small sales/planning/whatever cube consisting of 80 outlets, and 100 products, 12 months. Thats 96.000 values to be stored, each with a 100+ characters log line. That is 9MB logging for populating a small cube. That's maximum, I know, most cubes won't be populated so dense.

    The problem gets more obvious, if you think of updating a similar cube on a daily basis, e.g. with a "day" dimension instead od months. On my small test system, I will probably run out of disk space, when proceeding to more realistic tests. That's why I asked for logging customizability.
  • RE: logging customizable?

    Hello!
    The coming version has partially solved this issue by introducing additional log files (which theoretically can be schedule for archiving/deletion/rotation on a daily basis). The format however didn’t change. We wanted to stay loyal to human readable combination of CSV/XML.
    In your case you would get one additional cube archive file – containing log of all the actions (this file is not critical for server and can therefore be archived/removed). Current cube.log will then serve only for temporary changes between two commits. Once the server commits changes into cube.data, cube.log would be cleared, leaving you with fairly less used amount of space on the FS.
    Mit freundlichen Gruessen/ With kind Regards / Meilleures salutations

    Vladislav Malicevic
    VP Development & Support
    Jedox AG
  • RE: logging customizable?

    Is there any way to tranlate the cube name in the data directory to a more human understandable form?

    Eg in the example below the file U0FMRVMyMDA1.cube.log may equate to a palo cube called Test. Can you tell me how the data file naming convention works, since it TPI it was possible to determine which dimenions and cubes were which and therefore import them into RDMB's or copy the datafile from one model to another.

    Thanks

    The post was edited 1 time, last by Dave ().

  • RE: logging customizable?

    Hello!
    Take a look *.cube.xml. For example for cube Sales from Demo DB (U2FsZXM=.cube.xml):

    XML Source Code

    1. <?xml version="1.0" encoding="utf-8" ?>
    2. <Cube>
    3. Sales
    4. <Dimension>Products</Dimension>
    5. <Dimension>Regions</Dimension>
    6. <Dimension>Months</Dimension>
    7. <Dimension>Years</Dimension>
    8. <Dimension>Datatypes</Dimension>
    9. <Dimension>Measures</Dimension>
    10. </Cube>

    All the filenames are base64 encoded (e.g. Sales => U2FsZXM=).
    Mit freundlichen Gruessen/ With kind Regards / Meilleures salutations

    Vladislav Malicevic
    VP Development & Support
    Jedox AG