My Personal IT Infrastructure Knowledge Base

Thursday, May 25, 2017

WordPress - instalace

su -l [webmaster]
cd [web-document-root-directory]
fetch https://wordpress.org/latest.zip
unzip latest.zip
mv wordpress [site-name]

su -l root
cd /usr/local/etc/apache24/extra/
vi httpd-vhosts.conf

<VirtualHost *:80>
ServerAdmin david.pasek@gmail.com
DocumentRoot "/usr/home/[webmaster]/[site-name]/"
ServerName [site-name].dpasek.com
ServerAlias [site-name].dpasek.com
Options Indexes FollowSymLinks Includes
ErrorLog "/var/log/[site-name]-error.log"
CustomLog "/var/log/[site-name]-access_log" common
</VirtualHost>

apachectl restart

mysql -u root -p
CREATE DATABASE wp_[site-name] CHARACTER SET utf8 COLLATE utf8_bin;
grant all privileges on wp_[site-name].* to 'wp_[site-name]'@'localhost' identified by "pwd-[site-name]";

Wednesday, May 24, 2017

PowerCLI installation

Find-Module VMware.PowerCLI
Install-Module VMware.PowerCLI

Get-Module -ListAvailable VMware*

Update-Module VMware.PowerCLI

Friday, May 12, 2017

VCSA HA heartbeating

Q: What method is used for VCSA HA heartbeating (to validate that the primary VC is really not available)?

A:
There is a TCP hearbeat that happens every second between the nodes (initiated from the Active node). We monitor the active node via that heartbeat and ping. A failover is triggered when there are 3 lost heartbeats followed by 5 failed pings. Therefore, the node (or network) would need to be down for at least 8 seconds for a failover to be triggered.

The heartbeating technology that we use is based off of FDM (which is what vSphere HA uses) so it is a mature methodology that should work quite well.

Thursday, April 27, 2017

WordPress - mod_rewrite na staticke stranky

# BEGIN WordPress
<IfModule mod_rewrite.c>
RewriteEngine On
RewriteBase /
RewriteRule ^index\.php$ - [L]
RewriteCond %{REQUEST_FILENAME} !-f
RewriteCond %{REQUEST_FILENAME} !-d
RewriteRule . /index.php [L]
</IfModule>

# END WordPress
# BEGIN EWWWIO

# END EWWWIO

Friday, April 21, 2017

vSphere 6.5 - backup

I would like to follow up on the vSphere workshop we had on 9.3 and answer questions about vCenter 6.5 backup:

Q1: Is the backup single file? What is approximately a size?

A1: Backup is multiple files (screen1 attached) one per specific service. Approximate backup size might differ base on the number of components you are using (VUM, image builder and their data). During the backup process, it is calculated how much space it will approximately need (screen2), the portal in latest available version seems to be still unable to include amount of data from VUM and Imagebuilder, therefore 1.2GB expected by the tool differs by about 500GB from the real situation.

Q2: Best practice for backup of VCSA in HA mode

A2: VCSA in HA mode supports standard configuration backup through VCSA VAMI. In such case only configuration of the primary appliance is backed up. During the restore process VCSA is properly restored with HA mode being disabled -> afterwards HA mode should be re-enabled. In this case this was expected behavior as the VCSA VAMI backup is in-guest backup therefore it is not fully aware of the configuration of the other VCSA nodes (like Image level backup would be).

More information can be found: http://pubs.vmware.com/vsphere-65/index.jsp?topic=%2Fcom.vmware.vsphere.install.doc%2FGUID-AFF34FA6-B7CF-4AE0-9C12-C674F160682C.html andhttp://pubs.vmware.com/vsphere-65/index.jsp?topic=%2Fcom.vmware.vsphere.install.doc%2FGUID-F02AF073-7CFD-45B2-ACC8-DE3B6ED28022.html.

Test observations

During the tests I noticed strange problem, which was so far identified as a bug. For initial placement of the Secondary and Witness appliance is not supported to choose SDRS Cluster. Further in the deployment you can choose specific datastores (can be part of the SDRS Cluster) and that should be supported configuration – but it is still not accepted and you are not allowed to proceed with the deployment (screen3). I’m currently working with the PM team and engineering to clarify the setup and resolve the problem.