23.7.7_3 breaks Internet connection

[sbellon]

And for Dual IP stack, I assume one IPv4 and one IPv6 gateway should be marked? Or does it only apply to the IPv4 one (if the IPv6 one is DHCPv6 and not PPPoE)?

[franco]

Yes, this pertains to both IP families.


Cheers,
Franco

[SvenTorsten]

FYI - In my configuration there are in  fact two gateways (one internal to access the DSL modem) and no IPv6.

[franco]

OK, that means it could prefer the internal one when it has to decide which one to use (because both are not marked upstream).


Cheers,
Franco

[article15]

Hello,

I had the same issue, almost 2 or 3 times lost pppoe after latest update in a day. also my default gateway already checked. And have only 1 gateway
attached error log.
help would appreciated
thnx

[franco]

This doesn't look related at all, attached kernel panic:

db:0:kdb.enter.default>  show pcpu
cpuid        = 2
dynamic pcpu = 0xfffffe0095bc69c0
curthread    = 0xfffffe00ef704900: pid 15917 tid 100461 critnest 2 "python3.9"
curpcb       = 0xfffffe00ef704e10
fpcurthread  = 0xfffffe00ef704900: pid 15917 "python3.9"
idlethread   = 0xfffffe0017905560: tid 100005 "idle: cpu2"
self         = 0xffffffff82c12000
curpmap      = 0xfffffe00ef6dd518
tssp         = 0xffffffff82c12384
rsp0         = 0xfffffe00ed6f0000
kcr3         = 0x2d71b1000
ucr3         = 0x2d7007000
scr3         = 0x2d7007000
gs32p        = 0xffffffff82c12404
ldt          = 0xffffffff82c12444
tss          = 0xffffffff82c12434
curvnet      = 0
db:0:kdb.enter.default>  bt
Tracing pid 15917 tid 100461 td 0xfffffe00ef704900
kdb_enter() at kdb_enter+0x37/frame 0xfffffe00ed6ef810
vpanic() at vpanic+0x182/frame 0xfffffe00ed6ef860
panic() at panic+0x43/frame 0xfffffe00ed6ef8c0
trap_fatal() at trap_fatal+0x387/frame 0xfffffe00ed6ef920
trap_pfault() at trap_pfault+0x4f/frame 0xfffffe00ed6ef980
calltrap() at calltrap+0x8/frame 0xfffffe00ed6ef980
--- trap 0xc, rip = 0xffffffff80d5310a, rsp = 0xfffffe00ed6efa50, rbp = 0xfffffe00ed6efb10 ---
cache_fplookup() at cache_fplookup+0x20a/frame 0xfffffe00ed6efb10
namei() at namei+0x124/frame 0xfffffe00ed6efbc0
kern_statat() at kern_statat+0xf6/frame 0xfffffe00ed6efd00
sys_fstatat() at sys_fstatat+0x2f/frame 0xfffffe00ed6efe00
amd64_syscall() at amd64_syscall+0x10c/frame 0xfffffe00ed6eff30
fast_syscall_common() at fast_syscall_common+0xf8/frame 0xfffffe00ed6eff30
--- syscall (552, FreeBSD ELF64, fstatat), rip = 0x825595b1a, rsp = 0x8209db9d8, rbp = 0x8209dbb00 ---

[genesysguy]

So if not running PPOE at all, and I have 2 gateways (dual WAN provider), should I be checking upstream gateway at all or on both or just one?

[franco]

Pick the preferred one or none at all if the system should chose one for you.

Mark multiple as upstream if you use default gateway switching so it can use these to change between lines during downtimes (but it also requires the gateway monitor).


Cheers,
Franco

[article15]

What can i do about it?
thnx

This doesn't look related at all, attached kernel panic:

db:0:kdb.enter.default>  show pcpu
cpuid        = 2
dynamic pcpu = 0xfffffe0095bc69c0
curthread    = 0xfffffe00ef704900: pid 15917 tid 100461 critnest 2 "python3.9"
curpcb       = 0xfffffe00ef704e10
fpcurthread  = 0xfffffe00ef704900: pid 15917 "python3.9"
idlethread   = 0xfffffe0017905560: tid 100005 "idle: cpu2"
self         = 0xffffffff82c12000
curpmap      = 0xfffffe00ef6dd518
tssp         = 0xffffffff82c12384
rsp0         = 0xfffffe00ed6f0000
kcr3         = 0x2d71b1000
ucr3         = 0x2d7007000
scr3         = 0x2d7007000
gs32p        = 0xffffffff82c12404
ldt          = 0xffffffff82c12444
tss          = 0xffffffff82c12434
curvnet      = 0
db:0:kdb.enter.default>  bt
Tracing pid 15917 tid 100461 td 0xfffffe00ef704900
kdb_enter() at kdb_enter+0x37/frame 0xfffffe00ed6ef810
vpanic() at vpanic+0x182/frame 0xfffffe00ed6ef860
panic() at panic+0x43/frame 0xfffffe00ed6ef8c0
trap_fatal() at trap_fatal+0x387/frame 0xfffffe00ed6ef920
trap_pfault() at trap_pfault+0x4f/frame 0xfffffe00ed6ef980
calltrap() at calltrap+0x8/frame 0xfffffe00ed6ef980
--- trap 0xc, rip = 0xffffffff80d5310a, rsp = 0xfffffe00ed6efa50, rbp = 0xfffffe00ed6efb10 ---
cache_fplookup() at cache_fplookup+0x20a/frame 0xfffffe00ed6efb10
namei() at namei+0x124/frame 0xfffffe00ed6efbc0
kern_statat() at kern_statat+0xf6/frame 0xfffffe00ed6efd00
sys_fstatat() at sys_fstatat+0x2f/frame 0xfffffe00ed6efe00
amd64_syscall() at amd64_syscall+0x10c/frame 0xfffffe00ed6eff30
fast_syscall_common() at fast_syscall_common+0xf8/frame 0xfffffe00ed6eff30
--- syscall (552, FreeBSD ELF64, fstatat), rip = 0x825595b1a, rsp = 0x8209db9d8, rbp = 0x8209dbb00 ---

[franco]

Same as this apparently... https://forum.netgate.com/topic/182617/crash-reporter

Problem is the stack trace doesn't tell us much. Worse is that FreeBSD 14 appears to crash in the same way indicating that this is not fixed...


Cheers,
Franco

[Mks]

Dear all,

may you are interested in.

I faced the same issue "No internet connection" after update.

Even though I've only one gateway I had to check the "Upstream Gateway" checkbox, afterwards everything worked again.
Next issue I had was during the next reboot. Again the same issue and I had to "uncheck and check" the same checkbox again to make it work again.

br

[lar.hed]

Well I had (and have) internet connection (I could for example web surf to 1.1.1.1), but I did not have DNS name resolution. So I think there is at least two different issues here...

[franco]

> Next issue I had was during the next reboot. Again the same issue and I had to "uncheck and check" the same checkbox again to make it work again.

Misconfigured default gateway switching when using a single gateway and monitoring it? If it's down it will not be used.


Cheers,
Franco

[Mks]

Hi franco.

May I explain it a bit more in detail:

Before update:

• 2 Gateways configured (1xPPPoE, 1xIPv6 Tunnel)
• 1 Gateway enabled (PPPoE) and active
• Upstream gateway - disabled (was never enabled since years)
• Disable Gateway Monitoring - enabled
• IPv6 Gateway disabled

After update:

• No internet access possible. Initially I thought it is an unbound issue with external name resolution, but when I tried to PING external IP on the sense I got
• "No route to host" message", which points me to the a possible issue with the default route.
• no "default route" after update
• after checking "Upstream Gateway" route was added and it worked again

After next reboot I had to disable "Upstream Gateway"

br
 

[erlendlk]

Hi, just upgraded from 23.7.5 to 23.7.8_1, and I also got problems reaching out to the world outside.

Gateway-wise my system is pretty simple. I have two WAN gateways, one for IPv4 and one for IPv6, both created automatically, if I remember correctly. No gateway switching settings (default). I ended up in this thread via https://forum.opnsense.org/index.php?topic=37007.0

After having fumbled around with the "Upstream Gateway" setting and done multiple reboots, I ended up with a little bit of extra explicit gateway config. Before, this element in the config was just

Code: [Select]

<gateways>
    <gateway_item/>
</gateways>

but now has become

Code: [Select]

<gateways>
  <gateway_item>
    <interface>wan</interface>
    <gateway>dynamic</gateway>
    <name>WAN_DHCP</name>
    <priority>254</priority>
    <weight>1</weight>
    <ipprotocol>inet</ipprotocol>
    <descr>Interface WAN_DHCP Gateway</descr>
    <monitor_disable>1</monitor_disable>
  </gateway_item>
  <gateway_item>
    <interface>wan</interface>
    <gateway>dynamic</gateway>
    <name>WAN_DHCP6</name>
    <priority>254</priority>
    <weight>1</weight>
    <ipprotocol>inet6</ipprotocol>
    <descr>Interface WAN_DHCP6 Gateway</descr>
    <monitor_disable>1</monitor_disable>
  </gateway_item>
</gateways>

Not entirely sure where/why things stop, but after each reboot of OPNsense, I have to take action to get internet working again. It seems like pretty much every change that enables me to apply the gateway config brings internet/gateway(s) back, regardless of the "Upstream Gateway" setting.