If it's just nix based systems, Prometheus is super easy to set up. Prometheus, node exporter, and alert manager are really easy to automate and get going.

https://gitlab.com/hooksie1/ansible-prometheus

https://gitlab.com/hooksie1/node-exporter

I made these roles so all you should need to do is create a playbook to plug in the variables.

@black3dynamite said in Intranet suggestions....:

@stacksofplates said in Intranet suggestions....:

I'll recommend Drupal. It's the most flexible and you can build anything. It's great for very simple sites or really complex ones.

The farthest that I ever got with Drupal was installing it.

Pantheon makes playing around much easier. Anything I build I start on Pantheon and then move it to wherever.

@nashbrydges said in Intranet suggestions....:

Any of these options recommended for a multi-client scenario? Would need authentication so each client only accesses their own documentation.

Yes drupal can do that. Wiki.js can have roles and users that have access to specific areas but I've noticed if you search for something, the search show up in the bar from areas they don't have access to. They can't get there, but the titles and such show up.

@tim_g said in Fedora 27 Update History; surprise BIOS update:

@stacksofplates said in Fedora 27 Update History; surprise BIOS update:

@tim_g said in Fedora 27 Update History; surprise BIOS update:

@aidan_walsh said in Fedora 27 Update History; surprise BIOS update:

I had a microcode update applied by Korora 26 during the week, perhaps that?

Check dnf history?

Yeah dnf history shows my command/install history. But I don't know how to get it to show everything that was updated.

You can do

dnf history info <history number>

I've tried that but it doesn't show the whole line... it gets cut off so it doesn't help at all.

huh? There should be a lot more than one line.

Ya I also pretty much use Virt-Manager/Virsh. I have a bare KVM server and an OpenStack box. I pretty much use Ansible/Terraform to spin up new instances on OpenStack.

If nothing else, get off of 7.3 and move to 7.4. You can check the battery from the cli by typing acpi. If that works, it's probably whatever KDE is using for measuring the battery doesn't work with your system.

@wirestyle22 said in Installing Wiki.js on CentOS 7:

@tim_g said in Installing Wiki.js on CentOS 7:

@zachary715 said in Installing Wiki.js on CentOS 7:

@scottalanmiller said in Installing Wiki.js on CentOS 7:

@zachary715 said in Installing Wiki.js on CentOS 7:

@jaredbusch said in Installing Wiki.js on CentOS 7:

@scottalanmiller said in Installing Wiki.js on CentOS 7:

@jaredbusch said in Installing Wiki.js on CentOS 7:

Why are people using CentOS for This? Nothing on their website says it has to be on this.

Anyone using CentOS other than me? I'm using it because it is our existing standard platform for NodeJS deployments and we aren't moving from it anytime soon. I was asked to document my process, so I did. It should be even simpler to do it on Fedora, that's just not where I'm deploying right now.

Yes, one other in another thread was talking about CentOS for some reason. While someone else said they used Ubuntu I believe.

I used Ubuntu because it's what their documentation recommended.

Sort of, @wirestyle22 and I dug into that and while the docs say that, it's also pretty clear that they just don't update the docs, sadly. And they only document an install of Debian. They are all over the place. They don't exactly recommend old OSes like Ubuntu 16.04 and Windows 2012 R2, but point out that they are "more tested", which should be obvious since they are old. But that's not the same as recommended.

"More tested" doesn't mean "more stable", already broken down old things are often better tested than new, reliable ones. A barely functional 1975 Pinto is "better tested" than a brand new BMW 335i, but the new BMW is probably less likely to leave you stranded.

They word things a bit funny there and between that and not updating well, it can lead you in some weird directions. Not that Ubuntu is bad, but I don't think that they intend to recommend it in any way.

Wiki.js runs on pretty much any platform that supports the requirements below. However, the following environments are recommended and more thoroughly tested:

Ubuntu Server 16.04 LTS
Windows Server 2012 R2

I did not interpret that to mean that the OS itself was more tested, but rather that had tested Wiki.js on these OSes more so than others. They make it clear it can run on any system though. I just chose to go with what they had claimed to have tested for and it's worked out for me thus far.

To me, it seems better to stay away from Wiki.js until they take it more seriously.

I don't really see any additional benefits to using it versus other solutions that seem to work better and offer more.

It does seem like they are lax on testing

It's node so I don't think you will see much changing between platforms. VMware is running it on PhotonOS. I've run it on CentOS and Fedora and haven't had any issues. Small projects like this usually don't have time to test on a bunch of platforms. But again, since it's Node it should be fine on anything.

@scottalanmiller said in Centos 7.3 Installed:

@stacksofplates said in Centos 7.3 Installed:

@lakshmana said in Centos 7.3 Installed:

@stacksofplates for acpi "command not found" is the error

You'll have to install it.

yum install acpi

The package has one more letter in the name. Should be this...

yum install acpid

Then you can fire it up like this...

systemctl start acpid
systemctl enable acpid

ah my bad, posted in a hurry. Sorry about that.

@scottalanmiller said in Centos 7.3 Wifi Issue:

@black3dynamite said in Centos 7.3 Wifi Issue:

Why are you so determined to make CentOS work on a laptop? If you want to study for Red Hat exam, create a VM instead. CentOS hardly support hardware the requires propriety drivers.

Even when willing to use Fedora, I can't always make it work. My big laptop has to use Ubuntu. Nothing else can install on it, not even things based on Ubuntu like Mint!

Not to sidetrack this but did you try setting nouveau.modeset=no on your boot options? My XPS 15 has an NVIDIA something in it and Fedora wouldn't even boot normally (couldn't even run the installer). If you add that to your boot line it just uses the Intel stuff and you can then add bumblebee or whatever the NVIDIA junk is called later.

@lakshmana said in Centos 7.3 Wifi Issue:

@stacksofplates said in Centos 7.3 Wifi Issue:

@scottalanmiller said in Centos 7.3 Wifi Issue:

@black3dynamite said in Centos 7.3 Wifi Issue:

Why are you so determined to make CentOS work on a laptop? If you want to study for Red Hat exam, create a VM instead. CentOS hardly support hardware the requires propriety drivers.

Even when willing to use Fedora, I can't always make it work. My big laptop has to use Ubuntu. Nothing else can install on it, not even things based on Ubuntu like Mint!

Not to sidetrack this but did you try setting nouveau.modeset=no on your boot options? My XPS 15 has an NVIDIA something in it and Fedora wouldn't even boot normally (couldn't even run the installer). If you add that to your boot line it just uses the Intel stuff and you can then add bumblebee or whatever the NVIDIA junk is called later.

May I know the location to use this ?

That question was for @scottalanmiller about his ASUS laptop.

@tim_g said in Create NFS file share on Hyper-V Server 2016:

@scottalanmiller said in Create NFS file share on Hyper-V Server 2016:

@tim_g said in Create NFS file share on Hyper-V Server 2016:

Possible solutions:

• Create a simple fileserver VM on HV06, serving an NFS share that is located on the MD1000.
  • Because it's on a VM, the NFS share will be inside of a .VHDX file.
    • This makes restorations more time consuming or complex, because first the .VHDX would need to be restored in order to restore the NFS share inside of it... and then the backup data in the NFS share could finally then be recovered.

This is the part that is consistently incorrect. You do not need to restore the VHDX first.

Okay, so how do I get the data off of the VHDX, which is on a Tape, without restoring the VHDX?

One way is again to mount the VHD on the host and then pull the files out that way. Then it has nothing to do with NFS. Just mounting a drive and reading data.

@lakshmana said in Timeout in .sh file:

@thwr said in Timeout in .sh file:

I would always use curl / wget / any other CLI tool before I would use Chromium (Chromium != Chrome) for a simple file download. It's a monster in terms of resource consumption.

I tried wget where credentials is asking and in curl the file is not downloading where the test file created inside that file the link is there to download the file.So only using Chrome here

Show us the command you ran. I have a feeling you’re either not following a redirect or your output option is not set correctly.

I’ve used Asana and Nozbe. Both seemed fine, but they are SaaS. I don’t think they have self hosted options.

OpenProject looks interesting https://www.openproject.org/

Another option that’s kind of out of the box is to use GitLab. The issue tracking is pretty nice and with a subscription you can do burn down charts and some other cool stuff with your project milestones.

telnet

Holy crap

I just wanted to chime in and say sorry you have to use VNC.

@scottalanmiller said in I hope Wiki.js does not fail:

@stacksofplates said in I hope Wiki.js does not fail:

@scottalanmiller said in I hope Wiki.js does not fail:

@jaredbusch said in I hope Wiki.js does not fail:

@scottalanmiller said in I hope Wiki.js does not fail:

They've stopped work on the wiki.js 1.x series completely. ugh.

Reading the milestones for it, I’ll give them the chance.

It's worth following. But looking at their focus being on huge breaking changes that don't matter but not getting what they have working yet (no tables, can't save changes) doesn't bode well for how they approach the project. Like you said, just one guy. If he starts flailing around doing the "next cool thing" instead of getting it solid, that's a problem. He had a pretty solid thing already, had he just put a tiny bit more effort into it, it would be way ahead of where it is.

You can do tables. You just have to manually make them. It’s just the button doesn’t do anything.

What's the format? I tried that and it failed.

The normal markdown format, I’ll have to look. I think it’s something like:

| | |
————
| | |

That doesn’t format at all correctly. I’ll look when I get home

@gjacobse said in Remmina VNC: Unknown Authentication scheme:

@stacksofplates said in Remmina VNC: Unknown Authentication scheme:

I just wanted to chime in and say sorry you have to use VNC.

Well,.. I don't have to use VNC. But VNC is native to the rPi distro installed. I could install something like x2goserver, but even that has given me some trouble.

I’d just use X11 forwarding.

@jaredbusch said in Fedora 27 Filesystem root full:

so /usr and /home?

Ya. Is there another user on your machine?

@jaredbusch said in Dynamic DNS with CloudFlare:

This is interesting.
But what about jsut using a CNAME to your DynDNS name? That is how I handle it.

I have jared.bundystl.com as a CNAME on CloudFlare for sorvani.mooo.com from afraid.org. My router keeps afraid.org up to date.

I just don’t have to do accounts with anything else, that’s all. This just makes it a one stop shop. I actually stopped running it on my ERL and run it on one of the servers here. It’s been running since I did this and have never had any issues.

@aaronstuder said in KVM Templates:

@stacksofplates building templates using virt-builder

Oh. It's pretty simple. virt-builder --list gives you a list of images you can use.

Then just create your template with virt-builder <os> --format <format> imagename.format

Then you can use virt-customize to inject packages, create users, set passwords, etc. Then just clone that for your template.