Updated from .32 to 1.1.0 to 1.1.2 to 1.2.0 - User Login Hist page blank



  • The User Login Hist link has not worked since my first update from .32
    lhsfog/fog/management/index.php?node=report&sub=user-track shows a blank page.

    The Apache error log shows the following:

    [CODE][Fri Aug 01 11:16:07 2014] [error] [client 192.168.11.17] PHP Warning: require_once(./lib/ReportMaker.class.php): failed to open stream: No such file or directory in /var/www/fog/management/reports/User Login Hist.php on line 26, referer: http://lhsfog/fog/management/index.php?node=report
    [Fri Aug 01 11:16:07 2014] [error] [client 192.168.11.17] PHP Fatal error: require_once(): Failed opening required ‘./lib/ReportMaker.class.php’ (include_path=’.:/usr/share/php:/usr/share/pear’) in /var/www/fog/management/reports/User Login Hist.php on line 26, referer: http://lhsfog/fog/management/index.php?node=report[/CODE]

    On a whim, I copied the user login hist.php from .32 into /var/www/fog/management/reports/User Login Hist.php but that didn’t seem to do anything.

    Help??



  • Hello,

    I’ve upgraded to SVN 2906 and this problem is resolved.

    I now see a sensible page if I call up the login history.

    Thanks
    Alasdair


  • Senior Developer

    Would you be willing to upgrade to SVN?

    I have a feeling the issue you’re experiencing has already (albeit probably unknowingly) been fixed in the latest.

    ob_gzhandler is still done, but that shouldn’t be giving you a blank page.

    Does SuSE Linux have selinux with it as well? Just wondering.

    FTP is used during Image Uploads. It’s what moves the image from /images/dev/<macaddress> to /images/<imagename>.

    Being that it got stuck, it sounds, to me, like the password is not set properly, probably because of the upgrade.

    Why? I made an attempt to make this somewhat simplified, but I was not perfect. I’d say your easiest fix for that would be to set your own password for the linux fog user. This can be done with:
    [code]sudo passwd fog[/code]

    Enter your new password and then you will need to edit three locations.

    First (to allow other upgrades to work without issue – if my suspicions are correct) edit the:
    [code]/opt/fog/.fogsettings[/code] file and change the password= line to contain
    [code]password="<yourfogpasswordhere>"[/code]

    Then you will edit the node:
    Storage Management -> Search or List the relevant node -> Select the relevant node via the link provided -> Management Username and Management Password

    Then you will edit the TFTP_FTP password from:
    FOG Configuration->FOG Settings->TFTP Server->FOG_TFTP_FTP_PASSWORD

    You’ll replace the password fields to match that of the password you had set using the passwd command above.

    I don’t know if this will fix all or any of the issues, but I certainly hope it gives a good starting point to help troubleshoot.



  • Tom,

    I had misunderstood you. I thought you meant the fog logs. Attached are the apache logs. I have flushed the logs and then started a fog session ([url]http://localhost/fog/management[/url]), logged in and then I went directly to the images page and tried to list the images.

    the ob_gzhandler warning pops up with every transaction so I ignored that. Apart from that, I can’t see anything untoward.

    Can you explain your thoughts about the link between this problem and ftp? What does fog use ftp for? I understand that tftp and nfs are used and why but not ftp.

    Thanks
    Alasdair

    [url="/_imported_xf_attachments/1/1591_access_log.txt?:"]access_log.txt[/url][url="/_imported_xf_attachments/1/1592_error_log.txt?:"]error_log.txt[/url]


  • Senior Developer

    So I’m not sure where the problem lies or where the error logs are found on SuSE Linux. Maybe in /var/log/httpd or /var/log/apache

    This issues based on what you’re telling me appear related to ftp issues though.



  • Here is the images table from the db if it gives any clues

    [url="/_imported_xf_attachments/1/1590_images_table.txt?:"]images_table.txt[/url]



  • Hello Tom,

    Thanks for the fast reply.

    The memory suggestions in the thread seemed unlikely to me as I am currently only dealing with two images. Correct me if this assumption is wrong.

    This is what I did -
    allow PC (host) to create itself as a host on the server
    create an image for the host
    link the image to the host
    initiate a task to upload the image
    NB - this does not complete - stars appear on the host but the server doesn’t end the task. So, I move the image form dev/ to images/ with the assigned name of the image. I don’t change any rights.

    If I then check the image list - problem

    However, if I initiate a download, all seems to run smoothly.

    I have looked at the log files under /opt/fog but they don’t appear to have been touched since my last attempts a month ago.
    There are 4 logs - fogscheduler, fogreplicator, groupmanager, multicast
    Am I looking in the wrong place or have I managed to kill logging somewhere along the line?

    Best
    Alasdair


  • Senior Developer

    [quote=“Alasdair Hatfield, post: 40435, member: 24780”]Hello,

    Is there any advance on this problem?
    I’m running 1.2.0 on a SuSe platform.

    I haven’t discovered any problems with reports but I suffer the same problem as described above if I try to list the images or do a search.
    Clicking on ‘List All Images’ results in a blank page (page source is blank too - suggests that nothing is rendered)
    ‘New Search’ shows a page and if I enter a random string, I get ‘0 results found’ which is to be expected. If I start to enter the name of a valid image, I get ‘Internal Server Error’ as UpACreek does.

    Thanks in advance for any help.
    Alasdair[/quote]
    Did you try any of the suggestions as described throughout this thread? I don’t know what’s causing the issue on images but it sounds like either you have a ton of images or there’s an image or two that are not holding the proper I’d for whatever reason. On to that when you get the blank page what is in your error log?



  • Hello,

    Is there any advance on this problem?
    I’m running 1.2.0 on a SuSe platform.

    I haven’t discovered any problems with reports but I suffer the same problem as described above if I try to list the images or do a search.
    Clicking on ‘List All Images’ results in a blank page (page source is blank too - suggests that nothing is rendered)
    ‘New Search’ shows a page and if I enter a random string, I get ‘0 results found’ which is to be expected. If I start to enter the name of a valid image, I get ‘Internal Server Error’ as UpACreek does.

    Thanks in advance for any help.
    Alasdair



  • This release broke other functions as well.

    Trying to list all images, results in a blank page.
    Trying to search an image results in this:

    [url="/_imported_xf_attachments/1/1526_Uhoh.png?:"]Uhoh.png[/url]



  • I can’t even do a dates query. I click on either link, and it blanks out. I don’t get to a page where it will even begin to allow me to specify a query.


  • Senior Developer

    Or you simplifying the dates, or trying to grab all of them?



  • I’d love to report positive results… But… :(

    Both ‘Login History’ and ‘Hosts and Users’ result in a blank page after updating to SVN 2721. FOG_MEMORY_LIMIT is set to 1024.



  • Thank you for your continued work on this! Will test after Thanksgiving. gobble gobble

    Happy Thanksgiving everyone! :D


  • Senior Developer

    SVN 2705 released, should fix the memory problem for even the simplest of sortable data.

    Alright,

    SVN should contain the fix for this pressing problem. While I can’t fix huge lists and “Hosts and Users” will still be broken as it’s pulling all the data at once, it should work for all the other “timed” reports as my method was all screwed up. I had to add a new element to the find method to get it to work, but it seems to work properly now.

    Basically (and you all can slap me if you want) as this wasn’t added before, my mechanism for sorting out the proper datasets was to pull the entire set of entries. This was/is where the problem exists. I’ve added a new feature to allow me to adjust the find parameter’s Compare element. Because of this, I can now test anything I want so long as I know the syntax. It’s intended purpose, for now, is to simply use BETWEEN ‘$date1’ AND ‘$date2’ calls, but it has potential for much more down the road.

    Please test and let me know.



  • [quote=“Wolfbane8653, post: 38679, member: 3362”]How large is your userTracking Table in mysql?[/quote]

    Login records span from 2012 to now, and there are ~614,000 records just in the login history table.


  • Developer

    How large is your userTracking Table in mysql?



  • Bummer… still a blank page after going to Login History. :(
    Even tried changing the memory in the GUI to 1024, but still no dice.



  • I’ll give it a go. crosses fingers I’ll let you know.


  • Senior Developer

    Take a look at the current svn.

    You can update your memory_limit right in fog gui.

    I’m also taking an approach to use MYSQLI_USE_RESULT for the results page to allow, hopefully, a better size in the variable that’s stored, though it’s not likely to fix all issues.

    That said, you no longer need to, nor will it affect fog anymore, set the memory_limit in the php.ini file as it’s set in FOG and uses that value specifically. It defaults to 128 and will never go lower than that. Nice part is you don’t need to restart apache either for the changes to take effect.


Log in to reply
 

347
Online

39.3k
Users

11.0k
Topics

104.4k
Posts

Looks like your connection to FOG Project was lost, please wait while we try to reconnect.