KloxoMR slave server Nginx - error with skeleton.zip

[tochits]

Hello,

Whenever a client is created and a corresponding domain is made as well, there is an error with skeleton.zip that is unzipped. The document root directory does not have an index file created and the other standard subdirectories are not being created (e.g. public_html is missing, etc). This happens if the server is slave, running nginx, and the client is created on the slave.


sh /script/sysinfo

A. Kloxo-MR: 6.5.0.f-2015083001

B. OS: CentOS release 6.7 (Final) x86_64

C. Apps:
   1. MySQL: mysql55-5.5.46-1.ius.el6.x86_64
   2. PHP: php54-5.4.45-1.ius.el6.x86_64
   3. Httpd: --uninstalled--
   4. Lighttpd: --uninstalled--
   5. Nginx: nginx-1.9.6-1.el6.ngx.x86_64
   6. Qmail: qmail-toaster-1.03-1.3.55.mr.el6.x86_64
      - with: courier-imap-toaster-4.1.2-1.3.18.mr.el6.x86_64
   7. Dns: djbdns-1.05-17.4.mr.el6.x86_64

D. Php-type (for Httpd/proxy): php-fpm_event

E. Memory:
                total       used       free     shared    buffers     cached
   Mem:         31869        742      31127         26          0        193
   -/+ buffers/cache:        549      31320
   Swap:            0          0          0

[MRatWork]

Public_html is equal to 'default domain' document or the fist domain document root.

[tochits]

Reviving this thread since my original question was nver really answered.

I've tested master-slave setup again, now using v7.0. Again, the problem that I encounter is that the contents of skeleton.zip are not extracted to the client directory whenever a new client is created.

This problem is only if the client is hosted on the slave server. If the client is hosted on the master server, the skeleton.zip is properly extracted.

I've attached 2 screenshots to illustrate the problem. See that the "skeleton.zip*" is a 0-byte file (temporary file, even) for the slave server, while the contents are correctly extracted for the master server.

My version info:
sh /script/sysinfo
A. Control Panel:               
   - Kloxo-MR: 7.0.0.b-2016021403
   - Web: hiawatha-10.1.0-f.6.mr.el6.x86_64
   - PHP: php54s-5.4.45-2.w6 (cgi mode)
B. Plateform:
   - OS: CentOS release 6.7 (Final) x86_64
   - Hostname: kloxomr-master2.emc.com.ph
C. Services:
   1. MySQL: MariaDB-server-10.0.23-1.el6.x86_64
   2. PHP:
      - Branch: php56u-cli-5.6.18-1.ius.el6.x86_64
      - Multiple:
        * php52m-5.2.17-102.mr.el6
        * php53m-5.3.29-1.ius.el6
        * php54m-5.4.45-2.w6
        * php55m-5.5.31-1.ius.el6
        * php56m-5.6.17-1.ius.el6
        * php70m-7.0.3-1.w6
      - Used: php
   3. Httpd: httpd-2.2.31-1.mr.el6.x86_64
      - PHP Type: [unknown]
   4. Lighttpd: --uninstalled--
   5. Hiawatha: --unused--
   6. Nginx: --uninstalled--
   7. Cache: --uninstalled--
   8. Dns: --uninstalled--
   9. Qmail: qmail-toaster-1.03-1.3.55.mr.el6.x86_64
      - with: courier-imap-toaster-4.1.2-1.3.18.mr.el6.x86_64
D. Memory:
                total       used       free     shared    buffers     cached
   Mem:         15902        820      15082          1          0        126
   -/+ buffers/cache:        693      15208
   Swap:            0          0          0
E. Disk Space:
   Filesystem      Size  Used Avail Use% Mounted on
   /dev/simfs      845G   76G  770G   9% /


*** Process Time: 00:00:00:01.1036 (dd:hh:mm:ss:xxxxxx) ***

* Note: run 'sh /script/sysinfo -y' if you want run 'fix-service-list' also
        (importance after Kloxo-MR update)

[MRatWork]

Change php for panel with 'sh /script/set-kloxo-php fpm'.

[tochits]

This does not help.

The 'master-slave' setup in Kloxo MR is broken. A simple extraction of the skeleton zip file is always faulty.

I'm sticking with having a stand alone server moving forward.

[fossxplorer]

I've been testing slave for a couple of days and used quite some time to get it working.
Reading this thread confirm there is an issue related to zip (unzipping?) on a  master-slave setup.
Please see https://forum.mratwork.com/kloxo-mr-development/slave-change-password/msg52268/#msg52268

I don't it's a big fix, i hope @Mustafa gets some time to fix it. It's really useful when you want to scale your servers.
1 master Kloxo alone isn't enough.