Change of HDD
-
The entire config, interfaces and all.
-
It restores complete configuration, incl. packages. Start a new thread in the proper forum section regarding CARP.
-
Will do - sorry to divert a bit, wrongly so, from my original question!!
Thank you very much - it looks as if it will be far easier than I was looking at. I gave some help with the initial install, but got a genius to do most of it for me. Unfortunately, he passed away a few weeks ago - so, without him as a backup, I'm trying to make sure every possible problem is worked out in advance.
Just need to find a copy of 2.0.1 to install and then restore a backup to.
-
Just need to find a copy of 2.0.1 to install and then restore a backup to.
They are all on the mirrors…
http://files.atx.pfsense.org/mirror/downloads/old/
-
It restores complete configuration, incl. packages.
I seem to recall in the past a few examples of installed packages that exploded after being restored from config.xml. A raw image would be better than an XML backup, for sure, if they can spare the 5 minutes of downtime to image the drive.
-
Thanks for the link. Think I've got the right one to download now. I'll give it a try.
-
KOM, are you saying to clone the drive is the best option? Clonezilla? Or do you know of a better one?
-
If you are trying to find the best way to restore a system then yes, an image of the hard disk is the best way to go. Restore the image, reboot and done. No fiddling with checking that all packages and everything else are working after restoring XML. The only downside is the downtime. For imaging, I use Acronis B&R. I've used Clonezilla in the past but thought it was kind of clunky.
I'm not saying that Config.xml backup is bad; I also use it. But I take a weekly backup of my pfSense VM and I would rather restore that than do it from config.xml.
-
That roughly falls in with one of my plans/ideas.
That is to run pfSense from a USB stick (pen drive, whatever name you like!), rather than a 'normal' HDD, and to keep a couple of clones of it. That way, if there is a disk failure, we have a live version to simply plug in and reboot (quicker than opening a computer up to change HDD).
Failing that, on a serious PC 'death', we simply plug the USB disk into another PC, re-assign the ports to their new names and it running again quickly.
USB sticks are so cheap these days!!
-
But then you've lost any history you may have, logs, graphs, etc unless you're copying these sticks to each other on a regular basis.
-
usb-sticks will die quickly if you have packages on there that do a lot of writes. even just logging can wreck a usbdrive in a couple of months. (thats why there are specific memstick releases that limit writes).
-
@KOM:
A raw image would be better than an XML backup, for sure, if they can spare the 5 minutes of downtime to image the drive.
That won't even boot due to /etc/fstab. Note the "HDD to a USB drive" in the OP.
-
Forgot he wanted USB. I would never run that way myself. It doesn't feel natural. Plus I like my logs too much.
-
It is looking like using a USB HDD then - rather than a USB SSD.
Can it really knacker a pen drive in just a few weeks?
But might step into the upgrade from 2.0.1 to 2.2.2 by going to a USB stick first, on the basis that if there are errors, just I can just remove it an boot from the HDD.
-
I was looking at going to a Kingston 8Gb - it will do USB3, but only in a USB2 port
http://www.kingston.com/datasheets/DTSE9g2_en.pdf
It talks about a 5 year life?
