Netgate 3100 URL unknown
- 
 Hi Steve 
 have you done any further tests so far? Did you change the DNS accordingly?
- 
 I was able to review this again today. I replied on your ticket. I think we need to see exactly how you were configuring DNS in 22.05 that worked. Steve 
- 
 @stephenw10 said in Netgate 3100 URL unknown: I was able to review this again today. I replied on your ticket. I think we need to see exactly how you were configuring DNS in 22.05 that worked. Steve Hi Steve 
 I already pasted the screenshot with the "Google DNS workaround" and here as text:DNS Resolver: 
 Host Overrides
 4 4.8.8.in-addr.arpa 4.4.8.8
 8 8.8.8.in-addr.arpa 8.8.8.8
 Domain Overrides
 4.4.8.8.in-addr.arpa 127.0.0.1
 8.8.8.8.in-addr.arpa 127.0.0.1This was a workaround for anything above 22.01. In V21.x we had no workarounds needed to use our construct as we (want to) do now. My advise: disable the DNS parent in your second box and you will experience the problem. Our workaround for 22.01+ Update checks is just one point to a solution: The new dynamic Repos do not use the Proxy to resolve their names. Cheers 
 Michael
- 
 Hi Steve 
 my TAC Ticket vanished again (like 4month ago). Can you make it visible again?Cheers 
 Michael
- 
 Hi Michael, 
 I think your new ticket was merged into the old one which should have opened it again.I was able to get back to this today along with one of our developers. I think we have a handle on the issue. Or at least an issue that would prevent this working in your situation. We will probably more questions for you tomorrow on the ticket. Steve 
- 
 @stephenw10 well the new ticket does not include any info/replies of the old call and the call was vanished a week when I opened a new ticket with the new information what we collected so far. I find it quite disturbing when replies/tickets "vanish" as the logical checks are no longer plausible or understandable. Anyway let's see what comes next. 
 Cheers
 Michael
- 
 It should be on there I can see the replies internally. I replied on the ticket. We finally replicated the issue and it looks to be how Squid is handling the SRV requests/records. Digging continues. Steve 
 
