Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - Parsifal

Pages: [1] 2 3 ... 92  Next >
1
Status Notices / Re: Scheduled maintenance, 2019-10-13
« on: October 13, 2019, 10:54:17 AM »
So I slept through this window by accident. Whoops.

I'll schedule another for next weekend. There are some other changes I'd like to make anyway.

2
Status Notices / Scheduled maintenance, 2019-10-13
« on: October 10, 2019, 10:09:13 PM »
The homepage, forum and wiki will be going offline for about five minutes on 2019-10-13, between 09:00 and 09:15 UTC.

In your local time (if you are logged in and have your time offset correctly configured), this is:
October 13, 2019, 09:00:00 AM


The intent is to install security updates on the server which hosts the homepage, forum and wiki. These will be non-disruptive to functionality, as the server is running a stable OS release that gets critical fixes only.

3
Technology & Information / Re: My weekly OpenBSD time
« on: October 10, 2019, 06:48:29 AM »
Parsifal has taken the Star Citizen approach to development progress.

Almost as if I'm a student now...

4
Status Notices / Re: Scheduled maintenance, 2019-09-12
« on: September 12, 2019, 09:01:38 AM »
Maintenance complete.

5
Arts & Entertainment / Re: Gamergate!
« on: September 09, 2019, 09:57:45 AM »
But if it is true, isn't it more accurate to say that he's the only one responsible? He'd just be facing the natural fall out and consequences of his own actions. If people are caught the legal way doing horrible things and then commit suicide in jail (or while out on bail before trial or during the investigation), is it the legal system that drove that person to committing suicide? Or their own shame at being found out?

The important difference is that, when you get convicted with due process, you are given the opportunity to present your case to defend yourself. Even if you get convicted, you still have a voice and an opportunity to have both sides heard fairly.

There is no such opportunity with these SJW witch hunts.

6
Status Notices / Scheduled maintenance, 2019-09-12
« on: September 08, 2019, 03:34:23 PM »
The homepage, forum and wiki will be going offline for about five minutes on 2019-09-12, between 09:00 and 09:15 UTC.

In your local time (if you are logged in and have your time offset correctly configured), this is:
September 12, 2019, 09:00:00 AM


The server needs to be rebooted in order to complete the unblock for sending mail as described in the previous maintenance thread.

At the same time, we will install security updates on the server which hosts the homepage, forum and wiki. These will be non-disruptive to functionality, as the server is running a stable OS release that gets critical fixes only.

7
Status Notices / Re: Scheduled maintenance, 2019-09-07
« on: September 08, 2019, 03:58:39 AM »
Yes. I did that deliberately.

Do you have an actual point and/or problem, or are you just shitposting in Announcements for the hell of it?

8
Status Notices / Re: Scheduled maintenance, 2019-09-07
« on: September 07, 2019, 02:15:40 PM »
In the 7 years I have been on this site, I have never received a single e-mail from the forum.

That is because you have disabled e-mail notifications in your settings.

9
Status Notices / Re: Scheduled maintenance, 2019-09-07
« on: September 07, 2019, 12:25:04 PM »
I've already noticed one broken thing: Outbound e-mail is blocked by the new hosting provider. I've raised a ticket for this, but in the meantime you won't be able to receive e-mails from the forum.

For now, I've worked around this by proxying mail via the old server, so it should now work again. I don't expect a reply from the new hosting provider on this issue until Monday.

10
Status Notices / Re: Scheduled maintenance, 2019-09-07
« on: September 07, 2019, 09:53:38 AM »
I've already noticed one broken thing: Outbound e-mail is blocked by the new hosting provider. I've raised a ticket for this, but in the meantime you won't be able to receive e-mails from the forum.

11
Status Notices / Re: Scheduled maintenance, 2019-09-07
« on: September 07, 2019, 09:37:47 AM »
Maintenance complete. I still have some cleanup steps to take but everything should be operational now. Let me know if anything seems wonky.

12
Status Notices / Scheduled maintenance, 2019-09-07
« on: August 31, 2019, 11:19:35 PM »
The homepage, forum and wiki will be going offline for up to an hour on 2019-09-07, between 09:00 and 10:00 UTC.

In your local time (if you are logged in and have your time offset correctly configured), this is:
September 07, 2019, 09:00:00 AM


The server which hosts the homepage, forum and wiki is to be migrated to a new hosting provider at this time. No user-facing changes are expected.

At the same time, we will install security updates on the server which hosts the homepage, forum and wiki. These will be non-disruptive to functionality, as the server is running a stable OS release that gets critical fixes only.

13
Arts & Entertainment / Re: Star Citizen
« on: August 28, 2019, 10:38:55 AM »
Recently, Forbes had the following to say about star citizen

None of this is news.

14
Arts & Entertainment / Re: Star Citizen
« on: August 26, 2019, 08:03:56 PM »
I better go tell the developers not to even bother with it then. Rushy has already decided its fate. lols

I mean, have you even played Scam Shitizen? CIG's idea of "balance passes" is to alternate between making something ridiculously OP in one patch and utterly useless in the next. Meanwhile, they add new broken features faster than they can pretend to be fixing the existing ones.

You'd have to be delusional to assume this ship is going to be useful for any purpose except Chris Roberts's yacht fund.

15
Suggestions & Concerns / Re: Built in image hosting
« on: August 22, 2019, 11:03:48 AM »
I sort of agree - integrating another third party would probably just kick the can down the road - they could die at any time. On the other hand, if we want to store (potentially) every image someone uploads here, this could become a storage challenge over time.

Storage is cheap, and if it were to become a problem (which I doubt given our current volume of traffic), we could find ways to make it even cheaper with a little bit of work on SMF.

16
Suggestions & Concerns / Re: Built in image hosting
« on: August 22, 2019, 01:16:32 AM »
This would be easy enough to do ourselves without relying on a third-party service to host the images, which would be my preference if we implement something like this.

17
Status Notices / Re: Scheduled maintenance, 2019-08-11
« on: August 11, 2019, 09:05:12 AM »
Maintenance complete.

18
Status Notices / Scheduled maintenance, 2019-08-11
« on: August 09, 2019, 11:28:19 AM »
The homepage, forum and wiki will be going offline for about five minutes on 2019-08-11, between 09:00 and 09:15 UTC.

In your local time (if you are logged in and have your time offset correctly configured), this is:
August 11, 2019, 09:00:00 AM


The intent is to install security updates on the server which hosts the homepage, forum and wiki. These will be non-disruptive to functionality, as the server is running a stable OS release that gets critical fixes only.

19
Parsifal, thank you. I think modifying the dhcp server is within the scope of what can be changed/configured in our network environment. It's a university and we have very silo'd operations. I cannot directly modify any of the infrastructure, but I can request it to be done by someone else.

As far as simplifying os installs, I want to have an image with all the necessary apps and updates that can be installed litetouch or pxe. I want to be able to install unattended. Sounds like kickstart is the way to go.

In that case, I'd just follow Red Hat's documentation on setting this up:

https://access.redhat.com/documentation/en-us/red_hat_enterprise_linux/7/html/installation_guide/chap-installation-server-setup
https://access.redhat.com/documentation/en-us/red_hat_enterprise_linux/7/html/installation_guide/chap-booting-installer-x86#sect-booting-from-pxe-x86
https://access.redhat.com/documentation/en-us/red_hat_enterprise_linux/7/html/installation_guide/chap-kickstart-installations

This is the way that RHEL (and thus CentOS) is designed to work and the most travelled path. I wouldn't bother even looking at third-party tools unless you find that Red Hat's supported way can't do what you want for some reason.

I assumed, and planned on, having to setup a server for all of this, so that is not a problem.

I'm thinking "pull" is the way to go for app deployment, since clients can check in and get themselves up-to-date.

That seems reasonable to me. You just need to make sure your Kickstart script includes bootstrapping for whatever tool you choose (I'd probably go with Puppet here, but you can even have it just fetch and run Ansible locally if you want).

You should also ensure that you have some way to be notified of errors applying the configuration, so that you become aware of broken machines before someone notices and reports the problem.

20
What you want isn't possible because you have multiple conflicting requirements. You're going to have to decide which parts of what you want are important and which parts can be dropped.

I would like to be able to install the OS via litetouch, or PXE and have the ability to deploy applications centrally.
I have used DRBL, personally, but I'm not sure it will work in this environment without modifying the networking infrastructure - which I do not want to do.

Booting via PXE requires assistance from the DHCP server, which is part of the network infrastructure. You want to rely on the network infrastructure to do a job without needing to modify it. That isn't going to work very well.

If a new application needs to be added, I want to be able to push it out to all clients at once and not have to touch each machine.

There are, broadly speaking, three ways of accomplishing this:
  • Push-based configuration management (e.g. Ansible), which you've already rejected as it can skip hosts which are offline when run.
  • Pull-based configuration management (e.g. Puppet, Chef), which fetch their configuration from a central server whenever a machine is booted, or periodically. This requires you to maintain a central server.
  • Centralised installation of software on a network file server. Root on NFS appears to be possible on CentOS, but unsupported by Red Hat. In any case, this requires you to maintain a network file server.
Since you've already narrowed it down to the latter two options, you need to set up a server of some sort. Personally, I'd suggest going with the second option, as it's more in line with what Red Hat supports and doesn't break every machine if the server goes down for a few minutes.

Note that Red Hat seems to support Puppet, so that's probably your best bet for CentOS.

Also, I want to simplify the OS installs as well and not have to run through a bare-metal install for each new client.

I don't even know what you mean here, since you mentioned Kickstart previously. (Kickstart is the correct tool for automating CentOS installation, by the way, since it is what Red Hat supports.)

Are you trying to say that you want to avoid a manual installation for each new client?

Pages: [1] 2 3 ... 92  Next >