@teamits Agree.

@dominikhoffmann Virtual NICs under VMware Fusion. [image: 1615629358834-screenshot-2021-03-13-at-09.55.07.png] You could even have another virtual Mac sat the other side of a virtual pfSense install.

@mtsgsd said in Am I missing something here? Shuttle XPC no graphics, no serial port: only two pcie slots in this I5 shuttle So which Shuttle XPC computer do you actually have. My brief look suggested most have lots of connectivity options. http://global.shuttle.com/products/productsList?categoryId=105

@discarn8 Hey men, I had the same problem, exactly the same problem, my solution was reinstall pfsense usb via front. This allowed everything to work perfectly.

@bainsley said in new user and not sure about these log entries?: What is the firewall syncing with? the kernel? BSD? Which filter? The rules? The kernel includes capabilities what I would call a "IP stack". And it can act as a firewall. Both are needed so the FreeeBSD kernel can do it's firewall thing, and that's what pfSense is all about. Firewall rules are dynamically reloaded after many possible sirconstances, like processses restarting, alaiase changed, or some user GUI intercation. many others exist. So you will see this : @bainsley said in new user and not sure about these log entries?: check_reload_status 400 Reloading filter check_reload_status 400 Syncing firewall a lot. If you are not using IPSEC, this : @bainsley said in new user and not sure about these log entries?: php 414 rc.bootup The command '/usr/local/sbin/strongswanrc stop' returned exit code '1', the output was strongswan not running? (check /var/run/daemon-charon.pid) is just a message that pfSense stopped IPSEC - and it was already stopped. @bainsley said in new user and not sure about these log entries?: php 414 rc.bootup. Creating rrd update script This happens when pfSense finds a 'new' interface or object to make graphs for. The graph data is stored in a so called rrd file. A scripts that runs every xx seconds, populates the rrd file with data . So you can have, after a while : [image: 1615535267192-9cc0b9bc-bbd5-43ca-a771-a05bffd2cf8f-image.png]

https://forum.netgate.com/topic/161085/download-location-for-2-4-5-release-p1-amd64 -Rico

Ok I checked the upgrade log in /conf and there are a number of php errors. But I can't get any of it off the device. If I bring up cpsw1 with an IP I can ping things. But if I try and ssh off it it connect to the remote end and then hangs.

@adamw Yes, you always need to be on the correct version or else an install attempt will try to upgrade the package manager software, in this case to 21.02. https://redmine.pfsense.org/issues/10464

@p32spaceblaster I just tried this again but via the console/serial cable and the upgrade was successful.

@jimp Cool, thank you very much for the clarification. Cheers

@bigsy Thanks, I'll go over that guide. Upgrading sometimes can be like ds of that saying we jokingly used in the hospital "Well Mrs. So and So was doing OK until we started helping her."

I know this is old thread, but I wanted to provide information in case anyone runs into it also. Similar error to OP and similar resolution... (although this is with the latest release 2.5.0) I am setting up a refreshed pfSense box and using a quite old but little used Samsung 830 series 128GB SSD. It is a dell version that was removed from a laptop a good 7 years ago but didn't see much use except for test installations of Windows 10. During setup, I read the documentation which steered you towards ZFS, especially on UEFI modern gear, so I did that and left the default ZFS setup options selected. When executing the final Yes, it would attempt to do: kern.geom.part.mbr.enforce_chs: but then nothing, and eventually the screen would keep popping up with: ahcich1 timeout on slot x port x, a bunch of "SEND_FPDMA_QUEUED DATA SET MANAGEMENT" and just repeating timeouts. I then thought, let's try an old rusty spindle HDD and sure enough, that kern.geom command went right through and started copying files right away and no problems. Finger beginning to be pointed at the Samsung 830 (or its firmware) I then grabbed another SSD from my parts drawer, this time, an Intel X25-V 40GB (yeah.. old!), and tested another ZFS install and it too blew through without a drama. so similar resolution to the guy above, a different SSD seems to work, which makes me think its a firmware issue with the SSD perhaps (and 830 is very old series, so I'm not surprised to be honest...but i was trying to re-purpose some old gear so as to not waste resources!)

All fixed. It just needed a null modem adapter.

@teamits said in Need procedure to roll SG-3100 back to 2.4.5_1 from 21.02-p1(arm): I'd listen to Netgate over me. :) no one is forcing you unless you are a dedicated developer stay on the working version 2.4.5-p1

FYI, after a very positive experience with Netgate support , some more info on this. Indeed there is an issue with the upgrade process from 2.4.5-p1 to the 21.02 version. I would recommend staying at 2.4.5-p1 until image 21.02 gets published at Azure. We can't give any timeframes at this moment. Also: It depends on Azure, and 21.02 should have been already published. We expect it to appear in Azure Marketplace every day. So it seems: 21.02 should be available soon. The upgrade info in the FAQ is still valid, meaning: backup pfSense config and restore on a new instance. Regards, Kristof.

@jeremy11one said in How to upgrade to 2.5 instead of closed source 21?: @bmeeks Ok, that's what I thought too. I searched the update settings and the Advanced Settings pages, and Googled it for a while, but never found a fix. I'll be the guinea pig and try 2.5.0 on the SG-5100. Hopefully Netgate will add the automatic update feature for CE so I can remotely update again like all my CE brethren. Before I started, though, I would make sure I had a backup 2.4.5_p1 Factory Edition image or else a pfSense+ 21.02_1 image just in case the CE doesn't pan out. You will have to contact Netgate to get those images (the ones other than CE, of course). Another possibility is to put a second "hard disk" in the box via an m.2 card and write CE on that. You can configure the boot device in the BIOS I believe. That way you would preserve your existing factory image as a backup plan.

@ramses.sevilla You really should not do it, 2.5 has too much bugs at the moment, using it on a production cluster is not a good idea. And downgrading is much harder then updating ;)

@sgw I had exactly the same thing happen to me. GUI would not start but the device functioned correctly. I left it alone, ordered another SG-1100 and installed it, then reflashed the original one. So now I have a hot spare ready to go should I need it. As far as not waiting long enough, mine (updated via the GUI) sat for a long time on waiting for the GUI to reload, which it never did. That's why I power cycled it. Would it eventually have come up? Don't know. Anyhow, now I'm really gunshy about attempting another update, especially with all of the users reporting problems. I'm content to leave that green light flashing.