Messages

However, if its eating up disk space and not RAM, perhaps its worth to have a look what fills the disk. The first suspect I would look at is the /var/log directory for the log files. Often enough a sick process throws a lot of log entries which fills disk space very fast

Code Select Expand

opnsense # du /var/log

might give some insights. Moreover, if its filling up a particular log file, this file might then give hints which process to look at

Br br

... or run manually

/usr/libexec/locate.updatedb

....

Hi there,

well ... mine went somewhat different:

While after the dpkg update the update process needed to be started a second time as described here too, then the packages were downloaded and the upgrade started, suddenly a window popped up with a message 'Danger <some unspecific explanation, don't remember> ....' and then the GUI became unreponsive. Reload of the page gave 'address not reachable'. Fix was Option 11 'Reload all services' from console. After that, GUI dashboard showed version 25.7.6. Interesting was that also DHCPv6 server had to be restarted manually

Never seen that little danger window in any update before ...

Check reveals that update has been at the end successful.

Br br

Hz

Hier mal meine komplette Anleitung, abweichend von der Beschreibung in der Dokumentation habe ich mein Modem völlig dumm gemacht und überlasse der Opnsense das vlan7 tagging. Funzt seit Jahren sowohl VDSL (Draytek Vigor 167) als auch nahezu unverändert auf Glas (Telekom Modem2)

1. Modem:

VLAN-Tagging für ADSL und VDSL auf ,,Disable" stellen

2. Opnsense
anschließend wird die Konfiguration der Opnsense vorgenommen:

• Unter Interfaces -> Other Types -> VLAN ein neues VLAN anlegen:
• Parent Interface: (das Interface wählen, welches zum Modem führt, hier igb1)
• VLAN Tag: 7
• Description: VDSL
• Unter Interfaces -> Assignments beim WAN-Eintrag das VLAN-Interface aus (igb1_vlan7) wählen
• Speichern mit Save
• Unter Interfaces -> WAN:
• IPv4 Configuration Type: PPPoE
• IPv6 Configuration Type: DHCPv6
• PPPoE configuration / AnschlusskennungT-Online-NummerMitbenutzernummer@t-online.de, also AAAAAAAAAAAATTTTTTTTTTTT0001@t-online.de
• Passwort eingeben, das steht im T-online Brief
• Dial on Demand: CHECK
• Idle Timeout: 00 (Zwei Nullen, wenn man nur eine Null eingibt, klappt das zwar, die GUI vergisst diese aber beim Bearbeiten und man muss jedes mal eine Null eingeben. Wenn man zwei Nullen eingibt, geht das automatisch).
• DHCPv6 client configuration / Request only an IPv6 prefix: CHECK
• Prefix delegation size: 56
• Send IPv6 prefix hint: CHECK
• Use IPv4 connectivity: CHECK

Angeblich braucht man die 2 Nuller beim Idle Timeout nicht mehr, ich habe sie aber immer noch drin ....

Hoffe, das hilft ....

Br br

That's really interesting!
However, I have some difficulties to map your Server Update URL to the Update URL format given in [1].

The format as being defined in [1] look like
https://ipv4.api.hosting.ionos.com/dns/v1/dyndns?q=ZsWxJSLcX8Dpy40nu0fhycBXdnnTb0gM2yqKS3Xcmx4c3yA6EhZp3Sg0467MRvdn6dU4P0kQntSeskxpYsFMxXnrKmW8teSHSHc3eefgYu4tjb2veJH4tESXaTmW4tvNW6nrtSm6Vb6E7e2SpuqqjvFaqbBS4XGCS.
You may complement this with &ipv4=...&ipv6=... parameters which I could get working only so far from a FritzBox accordingly to have the desired result of an A AND an AAAA record in IONOS DNS for my domain. With all other routers, I could achieve only either or (either A or AAAA record). Even separate entries for ipv4 and ipv6 in dyndns config did not work.

Considering the format you mention though reads like there are some more parameters in your URL which leads to some questions:

1.) What means SECRET, API, from, PORTAL and from where are the required values here shown as xxxxx..... coming from?
2.) I assume that __MYIP__ is filled automatically by the check ip method ?!
3.) Is there a complement also for ipv6, eg __MYIPv6__ ?

Looking forward to a short reply

Br br

Und der Alternate Name/DNS Name des Client certificate stimmt auch mit dem Zielsystem/Aussteller überein?

@zickzack111

Mit welcher CA hast du denn die User Zertifikate erzeugt? Mit der CA von Opnsense?

Br br

Wenn ich das richtig verstehe, lebst du in einem Mehrfamilienhaus. Falls ja, dann spricht der Vermieter kräftig mit wie das gemacht wird. Folgendes Dokument fasst es ganz gut zusammen wie das aussieht und welche Möglichkeiten es gibt.

https://www.telekom.de/hilfe/downloads/glasfaser-technik.pdf

Es gibt ein sfp bei der Telekom von Zyxel, das recht gut auch in anderen Routern zum laufen gebracht werden kann (Zyxel PMG3000-D20B), auf https://telekomhilft.telekom.de/t5/Festnetz-Internet/FTTH-ohne-Modem-SFP-GPON-Modul/td-p/5998739 gibts da mehr Details. Aber wie gesagt, das hängt von der Inhouse Verteilung ab ...

Und ansonsten gibts hier diesen Thread dazu https://forum.opnsense.org/index.php?topic=36491.msg213593#msg213593

Br br

Hmmm, I can confirm that Sequoia 15.0  openssh runs fine with opnsense 24.7.4_1 /.5. No issues at all. both from my internal LAN and WLAN.

There is usually only one reason for such an error message - bad encryption or decryption. Aside of ssh implementation bugs (which is usually very unlikely) there is also the possibility of line or HW issue or brute force attacks... here, a detailed log would help ...

br br

OK, thanks this works also perfectly fine for me ...

Next time I will ask directly about a potential intention  ;D ;D ;D    its a shame to be second ....

br br

G'day,

I just noticed an exec rights issue with the 'locate' command in 24.7.1.

Executing as root eg 'locate newwanrc', it shows

Code Select Expand


locate: the locate database '/var/db/locate.database' is smaller than 256 bytes large.

To create a new database, please run the following command as root:

  /etc/periodic/weekly/310.locate


Executing then as root

  /etc/periodic/weekly/310.locate

It only outputs

Code Select Expand


Rebuilding locate database:
Must be root.

No database is created... Also the weekly update of this database is obviously not working

Before throwing something in Git: freebsd 14.1 issue or opnsense?

br br

Hi there,

I can also confirm that the -icmp2 kernel works fine here, both ipv4 and ipv6

br br

G'day,

Updated just to 24.7.1 which ran smooth as usual. Time to say a big thank you to all who made this happen - Great job again!!!

A (small) heads up I would like to bring to your attention which is the dyndns service resp. GUI behavior once the system reboots after update.

I am running ddclient with native backend and my dyndns update runs in two steps, one for ipv4 and one for ipv6. In between, my provider (dedyn.io) demands for a ~5 min pause and it is throttling the update requests to this interval. Already after the 24.7 (and now also 24.7.1) upgrade reboot, I am running then in a kind of (race) condition showing a very inconsistent picture about the state of the dyndns service. Reproducible, after the update reboot

• the Dashboard widget shows dyndns as 'not running', ps says it runs. Dashboard value is not changing without manual interaction
• the dyndns services page shows the correct new ipv6 address and the old ipv4 address
• the interface overview widget shows for WAN the correct new ipv4 address and ipv6 address

So, the state of internet connection reads somewhat confusing after the update ....

When manually restarting dyndns service on the dashboard service widget, nothing changes
When manually stopping dyndns and starting on the configuration page, the widget on the dashboard gets then the correct new ipv4 and ipv6 addresses; on the service configuration page, still the old ipv4 address is shown.
When manually restarting dyndns for a second time, everything is OK again.

So it takes in my case up to 10 min and two manual interactions until GUI values for WAN addresses are shown consistent to the reality.

I am aware that a fix of this is somewhat shaky as the service restart behavior is different from system to system
but perhaps over time, some optimization could be considered for this.

Br br

not to bother anyone with minor details and typos. But

Code Select Expand

# opnsense-patch -c plugins dd68ab68d


runs a bit better  ;)

br br

The DMZ and LAN Addresses are shown 2003:XXXX in your post. So: can you show the field to see whether prefix ID 0 and 1 are properly assigned to the ipv6 address on the DMZ and LAN interface?

What do you mean with IPv6 has not configured yet?

ipv6 on servers:  Assuming servers with Linux with ipv6 activated, you should then have at least a configuration on /etc/network/interfaces like

iface eth0 inet6 auto

Otherwise you won't get any ipv6 address on the servers