I completed migrating to a Proxmox base for my home infrastructure. Openmediavault it seems is not well suited as a virtualization and container platform. It’s good for a home NAS but I run into issues when I use the sharedfolder functionality for my VM and docker containers.
I wanted to move my WordPress blog to my Openmediavault server to consolidate my hardware. It use to be that I have this blog on a standalone FreeBSD box.
So what did I do to make this happen? Here are the high level steps that I did.
- Prepare the VM.
- Move the data.
- Run the App.
- Openmediavault is able to run Docker and also KVM.
I decided to use KVM because I wanted familiarity and I have not have a lot of experience with docker containers. However, I was also successful in migrating my other applications into docker containers ( more on this to follow.)
- Migrate the data.
I did a database export and copied over my /usr/local/etc and /usr/local/www folders. I imported the data using the mariadb commandline since I have not setup PhpMyAdmin yet. I encountered a snag on the import. It seems that my database will not import unless I did an individual export of the databases. Something to do with the database user conflicts. Another thing with the import is that I have to redo the database users. I had to drop the users, create user, grant privileges, and finally flush privileges.
The configs I had to check the users and groups as well as the root folder (www) and the wp-content.
- Run the app.
It would have been simpler if I just setup the server the same way. But, I decided to add an NGINX reverse proxy into the mix. I am able to make it work somewhat but ran into an issue when logging into the app.
this is the error I got:
fatal error: uncaught error: call to undefined function ctype_digit() in /usr/local/www/wordpress/wp-content/plugins/jetpack/vendor/automattic/jetpack-connection/legacy/class-jetpack-signature.php:244 stack trace: #0 /usr/local/www/wordpress/wp-content/plugins/jetpack/vendor/automattic/jetpack-connection/src/class-client.php(136): jetpack_signature->sign_request('i9a$fqo@pwxac&o…', 1578694208, 'ldozcjbmxi', 'vbvvpd3yni1/f2m…', 'post', 'https://jetpack…', 'query('jetpack.sso.req…') #3 /usr/local/www/wordpress/wp-content/plugins/jetpack/modules/sso.php(409): jetpack_sso::request_initial_nonce() #4 /usr/local/www/wordpress/wp-content/plugins/jetpack/modules/sso.php(392): jetpack_sso- in /usr/local/www/wordpress/wp-content/plugins/jetpack/vendor/automattic/jetpack-connection/legacy/class-jetpack-signature.php on line 244
I googled for days but I could never find any solution on the net. I finally decided to try to uninstall the Jetpack plugin. I had to go to wordpress.com and disconnect my blog. Once , I did that I was able to login. I then was able to update WordPress and re-enable the jetpack plugin. Booyah!!!
Installing QEMU VMs through Cockpit is not straightforward.
You still need additional configuration.
Here is the rough draft of what I did to get it working.
- setup bridge interface using standard linux setup.
- install brctl, dnsmasq, and ebtables.
- run virsh edit to modify the VNC port.
Here is how you run your own Poshmark share server.
- Install your favorite Linux/BSD distro on your fave VM host(Virtualbox for example)
- In my case, I have FreeBSD and I was able to use BHyve. * details on next post
- Install Python if you don’t have it already.
- Install Python Virtualenv (optional but highly recommended)
- Go into github and download the Poshmark share application.
- Install the needed requirements for the poshmark_sharing app.
- If you installed remotely, you can install XRDP or TightVNCServer
We are back up after we had a downtime after I upgraded FreeBSD. Not a total lost but I might have lost a few emails and blog posts here.
I will be diligently backing up from now on. Lesson not learned sad to say.
I am finally able to get a working LDAP working with my LetsEncrypt certificates.
The links that help me tremendously is Loga.us blog. I followed most of the instructions except for the replication piece. Also this Archlinux page gave me what i needed to fix the issue I was having.
Today I was doing a lot of upgrading.
I was able to purchase a new PC to get my pFsense install into the latest and greatest 2.4.3 version. My old router PC was an old Pentium 4 and pFSense needs a 64Bit CPU. I went ahead and got an old HP 8200 SFF which has a Core I5 CPU. At least, I am future proofing a bit when 2.5 comes up. I added a quad port NIC and I think I am set for now.
I had a small hiccup with the config restore because there was no easy way for one to tell which port is em0 or em1. Can’t really tell from looking at the connections and in the console. I guess one can do trial and error and move the connections and doing a ping. The only thing is after one reboots, the system will try to find an internet connection and one can’t really proceed. I found a bug report at the pFsense site which mentions a work around for this. Essentially, one has to disable the pkg source file so that one can skip to the console main screen. Took me about a good couple of hours troubleshooting.
Lastly, after I got my internet back and while I thought making a note of all of this in WP, I got another upgrade notification. All in a days work. Fun.
If one wants to make the startup code to work in the latest Eclipse. You need to install JAva 1.6 and change it on the Eclipse settings. It defaults to the latest one on my Mac.
new stuff day… lol