What does the status LED do when you apply power?

I would want to use an SSD if you are caching. The eMMC on the SG-3100 has a high write cycle life, it's not expected to fail within the life of the device. That's with the usual disk write rates pfSense has. We do not do anything special there to lower writes. However caching to it will increase disk writes significantly and the write life is finite so it must reduce the lifespan. If you're using Squid/Squidguard just for filtering sites you set it not to cache which would eliminate the issue. I would be more worried about using all available disk space with snort and squid installed to be honest. That could easily happen if you enable too many snort signatures or allow squid to log too much. Steve

Those default limit values in the sensors widget are not really valid for the SG-3100. You can adjust them. Temperatures in the 65-75C range are completely normal. Steve

The first place to check would be in the system log covering the period is stopped responding. Your description sounds like what happens if you exhaust the firewall state table, that would be logged if so. It could also be a php issue if you are always using the webgui when it fails. That too would be logged. Steve

Thanks, I'll do that one night this week and let you know. Sean

We sell power supplies separately, please contact our sales. sales@netgate.com

I don't have an SSD in my 3100 but the smartctl binary is there now, so it should be OK on 2.4.3. If you are comfortable running a snapshot, you can upgrade to a 2.4.3 snapshot now and test. If not, 2.4.3 should be out in a couple weeks.

Thanks for the info on the UPS… But its a bit pricey for home use for what it would be doing..  If I needed more run time would just spend the money on a bigger ups for my needs.  Then again the power outages in my area very very rare even with storms. My current run time handles the little flicker/brown outs get now and then..

I agree, this was suggested many times. I was told that this will be fixed in future website update.

ivor, No offense taken by your asking if I actually tried to read something besides a cereal box.  I get calls frequently reporting something does not work, or "I can't figure this out"…....... Each time I start with some standard inquiries.  Folks get hot when asked if their printer is on when they complain it does not work.... and,,, oh my,,,,, who turned it off?  I didn't do that.......... Thanks again for your assistance!

There's really no need for that. Today's SSD's are a lot more resilient. While what you're asking might be technically possible it's unsupported or recommended by us.

I can attest the value of the Gold subscription.  It has been worth every cent.

Please submit a ticket to our support and they will help out. You can just link this thread. Thanks! https://customercare.netgate.com/

@ivor: We will have exciting update soon for those who wish to upgrade from 2220. i'd love to hear it!

I had similar issues, the main thing is you need to fix the problem with the web-ui becoming unresponsive. To prevent this I went from using the resolver to the dns forwarder. I think the problem was caused by my netgear wifi "router" (used as an AP) flooding the SG-1000 - but moving to use the resolver fixed the issue. I also stopped using the netgear as a secondary DNS anywhere, not sure which one fixed it - but it's been rock solid for months now. Power cycling the sg-1000 (or any device) without a proper shutdown can cause filesystem corruption. I once had (not a netgate device, or a firewall) the apache config in the hosts file!!!! You maybe able to fix it by booting into single user mode (via the console) and running fsck (three times apparently). But definitely speak to support as suggested. Personally I'd save the config (if you can recover it) and re-flash.

As others have suggested, OP should contact our support and we'll get this sorted out. @gcu_greyarea: Had the same happen on a 2440. Netgate was so kind to replace the unit, but never received an explanation as to what caused the issue. My suspicion is C2000 bug, but I could be wrong …. This sounds familiar to me, didn't we talk about this? If I remember correctly we had a new guy working in warehouse and he closed the ticket after issuing RMA (which is correct step however it also stops notifications in case you reply). That's the feedback I received when I investigated this. But I may be confusing you with someone else, so please pm me your ticket number or email address you used to contact our support and I'll figure this out! Thanks!

Hello, Could you share status_output file? You can get that by browsing to [your_firewall_IP]/status.php.  (For example 192.168.1.1/status.php) That page can take longer to load than others as it contains a lot of data.  All of that is contained in a single compressed file, status_output.tgz, which is linked at the top of that page.

Nobody can recomend me a product?

Thanks for your input. What I worry about is that the POC was done with JavaScript but that does not mean other vectors are not available. I do get your point about executables not running on a firewall. Attack vectors are limited. Agreed.