New answers tagged domain-name-system
0
I know this post is ancient, but I found it today when querying "DKIM 2048 bit key with UltraDNS." My DNS team had attempted to split the key into two parts with quotes around them and a space between. That was causing UltraDNS to serve up 3 packets (the one in the center was empty) which caused inconsistent validation results.
What worked for me in the ...
0
OK... ...I'm only going to say this once... ...You cannot reach a private DNS through a NAT, unless the NAT is the DNS. If you have to port forward to get to a device, the device then has to be configured to respond on that port.
For instance, I send a DNS query on any port to any DNS on my internal network, I don't need a port number becuase they are not ...
4
If TXT record contains an SPF record then it definitely can break mail delivery.
0
Everything is possible. A TXT record is just text stored in the DNS, a human or a machine can read it and process it.
So, if you want, you can create a contact form or a script that will perform this check and look for a specific TXT value before doing something.
0
I assume by changing domain you will keep the old IP addresses of servers. When you migrate servers to new domain you can keep the legacy domain zone in DNS server. By doing this you have two FQDN for one ip address. If you want to use new domain as well; you should add all records to new domain in dns and add certificate in servers. You should also add host ...
0
If your old domain is going away, then I think one option would be to create a new DNS zone on the new Active Directory domain controllers with the name of your old domain. Then add the A records to the new zone. After your IIS servers are in the new domain, point their DNS to the new AD DNS servers and lookups to the old domain should resolve properly.
1
A Google search for bind dnssec returns a lot of relevant results. Two that look particularly promising are:
https://www.isc.org/dnssec/
which then leads to https://downloads.isc.org/isc/dnssec-guide/dnssec-guide.pdf
https://wiki.debian.org/DNSSEC%20Howto%20for%20BIND%209.9+
0
Even if you get the name resolution to work that way, there is no public endpoint for your traffic to enter your vNET and reach your server. If you can’t have a public IP on your server I would deploy a Azure Firewall that will NAT the traffic and point the DNS to the firewall public IP
https://www.petri.com/understanding-and-creating-nat-rules-in-azure-...
2
The reverse lookup zone is managed by the entity that manages that particular netblock, which almost certainly isn't you.
2
First of all, it's unclear what the actual goal is and it sounds like there could be an XY-problem kind of thing going on here.
Addressing the things you bring up, though, the name ns1.mydnsprovider.com. is not part of the example.com. zone (it cannot be, by definition), and unless the same nameserver happens to also be authoritative for the mydnsprovider....
0
It is a bit odd that Heroku will provide an https:// address but not allow you to redirect to it.
You might contact their customer support and ask how to setup the CNAME properly so that you can use the certificate they already have created for you.
Changing ...
www.my_domain | CNAME | www.my_domain.herokudns.com
is key to fixing this issue.
2
Since bind 9.2 (See the release notes here bind defaults to the setting
minimal-responses no-auth-recursive;
Try setting that to
minimal-responses no;
1
Assumming that tcp/udp connectivity to port 53 on the server is not blocked...
The default BIND named.conf is for a DNS to work "as a caching only nameserver (as a localhost DNS resolver only)"
To turn that into a public DNS server, you may apply some of these according to your needs:
list on any ip, by default it only listens in localhost
allow-query ...
0
After some headaches, I realized that the problem got fixed by rebooting the VM:
power_state:
delay: now
mode: reboot
message: Reboot to apply new hostname
timeout: 10
This is working, but conflicts with my Ansible provisioning executing after the machine is created. And it also delays everything. Furthermore, I want to understand why this ...
1
Instead, get a static IPv6 allocation, so the AAAA records do not need to be frequently updated. This can be provider independent space, but does not have to be.
If your ISP does not provide this on their business class service, get a better ISP.
1
Like what Sven said, using user specific hostnames (like customername.example.com) is a better solution. It is much simpler to implement and (depending on your registrar/name server provider) the DNS entries can be updated dynamically.
Also, subdomain hostnames are easier to implement solution as you don't need to manage the reverse proxy, just the DNS ...
5
Set up one primary server with Apache or Nginx as reverse proxy that forwards URLS like /server2 to server2.
That said, having specific host names for the customers (e.g. customername.example.com) is often a better solution - when useful, they could also use their domain name in this case.
0
"nslookup -query=srv _ldap._tcp.DOMAINNAME" worked for me, tried nslookup -type=srv _ldap._tcp.DOMAINNAME and didn't work.
Server 2008 R2
0
The DNS lookup for www.eastcottandburgess.com resolves to eastcott-burgess-deploy.herokuapp.com (it is a CNAME).
Using a globally distributed DNS checker (e.g. dnschecker.org) to lookup this host you will see that searches from different locations are returning different results in each location (some are actually quite similar). All of the results are IPs ...
0
I managed to resolve the issue by upgrading both the master and node pools to the latest version of Kubernetes. Unfortunately was I not able to pinpoint the exact issue.
1
often see DKIM configuration guides using default as the selector. Is this a special selector
It is not. There is nothing in RFC 4871 that says "default" as selector as a specific meaning.
Instead its section 4.1 introduces selectors as labels in the _domainkey zone that way:
To support multiple concurrent public keys per signing domain, the
key ...
42
It's brute force.
They have looked up the IP addresses of every domain name they can find, and then compiled the results into their own database.
0
Just figured it out. zone should be set to Realm:
samba-tool dns delete localhost site.example.pl site-ad.site.example.pl A 10.10.37.166 -U administrator
samba-tool dns delete localhost site.example.pl site.example.pl A 10.10.37.166 -U administrator
0
I had similar issue with KubeDNS. Din't find the root cause, but managed to fix by installing CoreDNS instead.
Since Kubernetes 1.11 CoreDNS is used by default as a cluster DNS server, so I highly recommend to migrate to from KubeDNS to CoreDNS.
Consider this doc if you decide to proceed -
https://kubernetes.io/docs/tasks/administer-cluster/coredns/#...
0
Refer to @Michael Hampton♦ 's answer, here is how I fix it.
Add a default server block (ssl), need to include ssl certificate paths to make it work, see here.
server {
listen 443 default_server;
listen [::]:443 default_server;
ssl_certificate /path/to/your/certificate
ssl_certificate_key /path/to/your/certificate_key
...
1
To my knowledge the QNAME trigger in RPZ has the same wildcard semantics as DNS in general. Ie, it's only when it's specifically the left-most label that is * that the asterisk is considered a wildcard.
With RPZ out of the picture, I don't believe there is any standardized interface that does what you ask for, however some DNS server implementations provide ...
4
Your description says that another domain name you don't own has been set to your IP address, and web visitors to that domain therefore see your site.
To fix this, you need to restore the default nginx server block which shipped with nginx (or in this case, Ubuntu's custom version of the default server block). This default server block does not serve ...
1
You can find the relevant configuration in Consul docs:
https://learn.hashicorp.com/consul/security-networking/forwarding
For instance, if you use dnsmasq you could add the following line to dnsmasq.conf
server=/consul/127.0.0.1#8600
It basically means that queries to the consul. domain and its subdomains
should be resolved by a resolver available at ...
0
I believe you can just use CNAME record to point your domain to AWS classic loadbalancer.
3
DNS is only responsible for resolving names to ip addresses. If you can access the camera via the name through HTTP/HTTPS and stream MPEG then it's not a DNS issue. DNS has resolved the name. If H.264 doesn't work then it's likely a firewall or configuration issue.
2
It sounds like everything is working as intended and that the confusion here stems from not knowing how DNS wildcards are supposed to work.
The wildcard matches any names below the point of wildcard record which are in branches that do not exist (keep in mind, DNS is a tree).
When you added the new *.co.uk.example.com record (doesn't particularly matter ...
1
This is typically handled by creating multiple TXT records. You can see _spf.google.com do it.
"v=spf1 include:_netblocks.google.com include:_netblocks2.google.com include:_netblocks3.google.com ~all"
In your case you would create new TXT records similar to...
spf1.example.com 300 IN TXT "v=spf1 include:mail.zendesk.com include:stspg-customer.com ...
2
You should not just point your hostname e.g. to the imap.google.com as there's TLS involved; the certificate doesn't match your domain!
While there is an SRV record defined for IMAP (RFC 6186, 3.2), client's don't globally support it, and you would need to use it on client's email domains, without your domain in between.
The most compatible option would be ...
1
For your scenario you typically end up with 4-6 DNS records.
Two A or AAAA resource records that point to the IPV4 (for instance 192.0.2.1) resp. IPv6 (for instance 2001:db8:10::1) ip-address of your webserver that will be accessible for visitors using the bare and www versions of your domain i.e. https://example.com and http://www.example.com.
One A or ...
0
Figured it out. It turns out all that was needed was for the other nameservers to know about the other nameservers, but as NS records (they were listed as A records, thinking that NS was reserved just for name servers for that particular SOA which is not the case).
1
I'm embarrassed to admit this, but in the end,the whole problem was due to a typo. The name of the DC is not actually dc1, it has a longer designation and the word server is abbreviated within it. I'm used to abbreviating "server" with "srv", my predecessor on the other hand used "svr". Clearly, the correct abbreviation (svr001dc) was thrown at me in ...
0
dc1 typically should would be resolvable from nslookup depending on some regkey values.
dc1.example.com should be resolvable.
First off confirm there is an actual A record for the domain controller in DNS.
After this do a full health check of AD as this CAN be a indicator of a bad domain migration and FSMO roles being orphaned but I doubt this is the ...
0
As per the documentation you need to whitelist accounts.google.com and *.googleapis.com.
Using dig at storage.googleapis.com:
$ dig storage.googleapis.com
I get this:
[...]
;; ANSWER SECTION:
storage.googleapis.com. 3448 IN CNAME storage.l.googleusercontent.com.
[...]
So you can see that storage.l.googleusercontent.com is an alias from ...
1
First, test connect from console:
sudo openvpn --verb 1 --config /path_to_conf/config.ovpn
If you have this error:
/etc/resolvconf/update.d/libc: Warning: /etc/resolv.conf is not a
symbolic link to /run/resolvconf/resolv.conf
then fix /etc/resolvconf/update.d/libc
before
DYNAMICRSLVCNFFILE="/run/resolvconf/resolv.conf"
after
#DYNAMICRSLVCNFFILE="/...
1
Solved!
Oh my god after like 6 hours of google & try and error i got it.
The zone (why ever) wasnt added to the named.conf by cwp ...
can be closed!
2
According to Heroku Dev Center Custom Domain Names for Apps there's a different pattern for TLS hostnames:
If you’re using the SSL Endpoint, note that your DNS Targets will be a
little different. For apps in the Common Runtime, the endpoint domain
name will have a name in the form of example-12345.ssl.herokudns.com.
While every app may be served on ...
2
Thank you for including your actual domain.
Your have configured the following sub domain delegation for cloud.xibis.net:
cloud.xibis.net. 60 IN NS dns.xibis.com.
And you have the correct A record for dns.xibis.com. as well:
dns.xibis.com. 60 IN A 104.47.165.245
And when you query your MX records directly on the authoritative name ...
1
First off, as a general observation, the name . does not refer to "the domain itself" but to the root of the DNS tree.
Specifically in the context of MX, though, the rdata 0 . has been defined to mean "null", as in "I do not want mail delivered".
No MX, on the other hand, has a completely different meaning; try delivering to the address directly associated ...
Top 50 recent answers are included
