ssossossosso

Forums

Covering OroCRM topics, including community updates and company announcements.  Subscribe

This topic contains 24 replies, has 5 voices, and was last updated by Dima Soroka Dima Soroka 1 year, 6 months ago.

  • Creator
    Topic
  • #16278
    bruwster
    bruwster
    Participant

    I’m assessing OroCRM for future use, on a feature basis is superb. But on a trial basis we’re having technical problems that have left our two installs (both on AWS) now showing 503 errors. They worked for a few days but on one of these I installed the MailChimp extension, got an Ajax error, then it goes 503. The other one when 503 without installing any extensions.

    My sysadmin (contractor not employee) says the following,
    “I have looked into the situation and see both instances are giving 503 for now. I have found one outofmemory error in system logs and restarted php-fpm and nginx services but it didn’t help. Server reboot didn’t help as well. The error messages in orocrm logs don’t make any sense as well. I consider OROCRM is too buggy and unpredictable from my brief experience working with it and would suggest to use another solution here.
    I could definitely reinstall it once more, but cannot guarantee anything after that.”

    Is he right, is OroCRM just too buggy? I’d like him to be proved wrong but…

Viewing 15 replies - 1 through 15 (of 24 total)
  • Author
    Replies
  • #16280
    Dima Soroka
    Dima Soroka
    Oro Core

    hi @bruwster

    Not sure what is root cause of you issue (we can take a look if you provide us access to your environment), but statement from your sysadmin is not informative at all (we will appreciate any additional information from web servers, php and oro log files). We have a lot of production instances running, if you find a bug – please feel free to report on GihHub or contribute fix.

    #16281
    bruwster
    bruwster
    Participant

    Any other thoughts from users out there?

    #16506

    Razva
    Participant

    It’s not buggy, it’s just “unpredictable”. They devs are using a specific setup which – obviously – works for them, and for all other users that use that *specific* config. If you try to use a kinda different approach…yeah…it’ll usually end with 500 and 503 errors.

    #16521
    Dima Soroka
    Dima Soroka
    Oro Core

    Hello @Razva

    Could you please be more specific, we would like to understand what is *specific* config which leads to 500 or 503 error and and fix the issue.

    #16527
    bruwster
    bruwster
    Participant

    Thanks for your comment Razva! We haven’t given up, guess its just a learning curve with a new system.

    #16533

    Razva
    Participant

    @Dima I’ve replied you to the other other topic. It’s basically a CentOS 7 Base Install. I’ve encountered these issues with 1.8, and usually the fix was to chmod 777 everything (which is a *very* bad idea).

    @bruwster the folks around here are VERY nice and usually reply in less than 24 hours. I had a hard time understanding the initial process, but once you do…it’s a go. If you cannot make it work please feel free to PM me and I can try to assist you (with no costs), as the other good guys from this community helped me at first. 🙂

    #16599
    Dima Soroka
    Dima Soroka
    Oro Core

    Thanks @Razva

    chmod 777 is definitely not a good fix, our installer checks and warns about directories that should be writable so if all checks passed application should work just fine after installation. We’ll double check if there are no missed checks.

    #16606
    Yevhen Shyshkin
    Yevhen Shyshkin
    Oro Core

    usually the fix was to chmod 777 everything (which is a *very* bad idea)

    There is a better way to do that, check this article – you should apply permissions to app/cache, app/logs, app/attachment and web directories, and if you not installed yet – to the parameters file (usually app/config/parameters.yml). Usually it should be done automatically during composer install or update using ScriptHandler call. But if you somehow avoided usage of composer – you can run that command manually:

    #16608

    Razva
    Participant

    I’ve just run this:

    I’ve also tried to manually apply chmod 777 to the folders:

    Then I removed the entire content from “cache/” and “logs/” and re-run both ideas.

    The error is still there. When accessing “/email/user-emails” I get “500 Internal Server Error”.

    Should I use Git not the tarball you’ve provided…?

    Thanks!

    #16609
    Yevhen Shyshkin
    Yevhen Shyshkin
    Oro Core

    So, all permissions are correct. Probably, issue is somewhere else, need to check PHP and web server logs.

    > Should I use Git not the tarball you’ve provided…?
    Generally – there is no difference, in case of git it will be easier to update your application.

    #16610

    Razva
    Participant

    As specified before, there’s really nothing in Apache’s error_logs. Checked the error_log generated by the vhost, and the general error_log. Nothing.

    In Oro’s prod.log I can see this: http://pastebin.com/ZynUxyVX

    #16625
    Yevhen Shyshkin
    Yevhen Shyshkin
    Oro Core

    Look’s like issue is here:

    Uncaught PHP Exception Twig_Error_Runtime: “An exception has been thrown during the rendering of a template (“You have requested a non-existent service “gadget.ro”.”)

    Can you tell me where “gadget.ro” is came from?

    #16626

    Razva
    Participant

    “some-address@gadget.ro” is a system mailbox, but I have nothing *just* named “gadget.ro”. Any hints?

    #16635
    Michael
    Michael
    Oro Core

    Have you edited parameters.yml or any other config files manually ?

    Can you please check for occurrences of @gadget.ro in yml files:
    find . -name '*.yml' -exec grep '@gadget.ro' {} \;

    #16637

    Razva
    Participant

    Sure, here you go:

    I cannot really see any result regarding “gadget.ro” in any file.

Viewing 15 replies - 1 through 15 (of 24 total)

You must be logged in to reply to this topic.