summaryrefslogtreecommitdiffstats
path: root/roles/openshift_node_dnsmasq
Commit message (Collapse)AuthorAgeFilesLines
* Use openshift.node.dns_ip as listening addressScott Dodson2017-08-141-1/+1
|
* Simplify generation of /etc/origin/node/resolv.confScott Dodson2017-07-261-11/+18
| | | | | | | | Don't attempt to copy /etc/resolv.conf whenever there's no watermark, just write out the list of nameservers that network manager tells us about like we do for origin-upstream-dns.conf Fixes https://bugzilla.redhat.com/show_bug.cgi?id=1474707
* Prevent the script to use default route ip as upstream nameserver.Steve Teuber2017-06-301-1/+2
|
* Use default ports for dnsmasq and node dnsScott Dodson2017-06-303-3/+4
|
* Run dns on the node and use that for dnsmasqScott Dodson2017-06-305-4/+24
|
* Push to the registry via dnsScott Dodson2017-06-181-0/+3
| | | | | | Configures OPENSHIFT_DEFAULT_REGISTRY=docker-registry.default.svc Adds 'cluster.local' to dns search on nodes via dispatcher script Adds '.svc' to NO_PROXY defaults
* Disable negative caching, set cache TTL to 1sSteve Kuznetsov2017-06-131-0/+2
| | | | | | | | | | | | When running headless services as pods on OpenShift, no ClusterIP is assigned. In these cases, peer pods are relying on DNS to locate the service endpoints. When a pod is deleted or another trigger causes the endpoint to change, the OpenShift DNS is updated immediately. However, dnsmasq has a default TTL of 30s, so the wrong response is returned on name resolution. Removing negative caching and turning the TTL to a very short 1s should resolve the issue. Signed-off-by: Steve Kuznetsov <skuznets@redhat.com>
* YAML LintingRussell Teague2016-12-121-1/+1
| | | | | * Added checks to make ci for yaml linting * Modified y(a)ml files to pass lint checks
* node_dnsmasq - restart dnsmasq if it's not currently runningScott Dodson2016-12-051-0/+4
| | | | | Fixes Bug 1401425 Fixes BZ1401425
* openshift_node_dnsmasq - Remove strict-order option from dnsmasqScott Dodson2016-12-012-2/+0
| | | | | | | | | | | strict-order forces dnsmasq to iterate through nameservers in order. If one of the nameservers is down this will slow things down while dnsmasq waits for a timeout. Also, this option prevents dnsmasq from querying other nameservers if the first one returns a negative result. While I think it's odd to have a nameserver that returns negative results for a query that another returns positive results for this does seem to fix the issue in testing. Fixes Bug 1399577
* Merge pull request #2853 from alexcern/dhcpv6Scott Dodson2016-11-291-1/+1
|\ | | | | Set nameservers on DHCPv6 event
| * Set nameservers on DHCPv6 eventAlex Lossent2016-11-231-1/+1
| | | | | | | | | | | | A dhcp6-change event may happen on nodes running dual stack IPv4/IPv6 and DHCP, even if Openshift itself doesn't use IPv6. /etc/resolv.conf needs be adjusted as well in this case.
* | Refactored to use Ansible systemd moduleRussell Teague2016-11-283-9/+9
|/ | | | | | * Ansible systemd module used in place of service module * Refactored command tasks which are no longer necessary * Applying rules from openshift-ansible Best Practices Guide
* Make the role work on F25 CloudMichael Scherer2016-11-221-0/+1
| | | | | | On F24 and earlier, systemctl show always returned 0. On F25, it return 1 when a service do not exist, and thus the role fail on Fedora 25 cloud edition.
* Merge pull request #2818 from mtnbikenc/package-refactorScott Dodson2016-11-211-1/+1
|\ | | | | Refactor to use Ansible package module
| * Refactor to use Ansible package moduleRussell Teague2016-11-171-1/+1
| | | | | | | | | | The Ansible package module will call the correct package manager for the underlying OS.
* | Create the file in two passes, atomicly copy it overScott Dodson2016-11-181-6/+8
| |
* | node_dnsmasq -- Set dnsmasq as our only nameserverScott Dodson2016-11-171-1/+4
|/
* Only restart dnsmasq if the DNS servers have changedTim Bielawa2016-11-011-6/+41
|
* Enable dnsmasq serviceScott Dodson2016-10-241-0/+6
|
* Merge pull request #2112 from sdodson/dnsmasq-additional-configScott Dodson2016-07-012-1/+17
|\ | | | | Add support for supplying a dnsmasq.conf file
| * Restart dnsmasq encase it was already runningScott Dodson2016-07-011-0/+1
| |
| * Add support for supplying a dnsmasq.conf fileScott Dodson2016-07-012-1/+16
| | | | | | | | | | | | Useful for POC environments where DNS may not actually be setup yet. Make sure you don't define anything that would conflict with the default configuration. You have been warned.
* | Merge pull request #2028 from IshentRas/fix_dnsScott Dodson2016-06-301-2/+5
|\ \ | |/ |/| Improve 99-origin-dns.sh
| * - Prevent the script to override n number of the time the same nameserverIshentRas2016-06-131-2/+5
| | | | | | | | - Prevent the script to echo blank values from IP4_NAMESERVERS variable
* | NetworkManager service never changesTim Bielawa2016-06-221-5/+5
|/ | | | | * Don't ever report the NetworkManager service as changed when preparing to set the 'network_manager_active' fact
* Use grep to decide when to add our commentScott Dodson2016-05-261-1/+4
| | | | | | According to the man page for resolv.conf # needs to be in the first collumn in order to start a comment. While this didn't seem to affect glibc's resolver it did really confuse the kubelet which parses the host's resolv.conf
* Removed the echo line and replaced it with inline comment. To keep ↵Jonas Nordell2016-05-231-2/+1
| | | | 99-origin-dns.sh from adding a new line in /etc/resolv.conf everytime the NetworkManager dispatcher script is executed.
* Sync comments with origin prScott Dodson2016-05-121-4/+6
|
* Use IP4_NAMESERVERS rather than DHCP4_DOMAIN_NAME_SERVERSScott Dodson2016-05-121-3/+4
| | | | | Only modify resolv.conf if IP4_NAMESERVERS is defined, it's not safe to do so otherwise.
* listen on all interfacesScott Dodson2016-05-021-1/+0
|
* Make NetworkManager failure friendlierScott Dodson2016-04-291-1/+1
|
* Add openshift_node_dnsmasqScott Dodson2016-04-197-0/+117
generated by cgit v1.2.1 (git 2.18.0) at 2024-05-07 13:21:54 +0000