ansible/inventory/host_vars/db-datanommer01.stg.iad2.fedoraproject.org
Nils Philippsen dbbf94a411 ipa/client: configure global shell access and sudo
Almost global anyway, i.e. inside the VPN.

The ipa/client-based shell access and sudo rules are only effective for
staging right now, the respective playbook bits are masked out for prod.

- Assign Ansible host groups to IPA host groups, the latter don't care
  about 'stg' in the name and use dashes rather than underscores.
- Distill shell access groups from fas_client_groups in group and host
  vars.
- Let all `sysadmin-*` groups in the previous list run anything via sudo
  in the host group (except bastion & batcave).
- Remove `fas_client_groups` from staging host and group vars.
- Remove sudoers from staging host and group vars if only `sysadmin-*`
  groups have shell access.
- Set up `ipa_client_shell_groups` on bastion to be a super set of the
  same on batcave.

Newly created IPA host groups:
- autosign
- badges
- basset
- bastion
- batcave
- blockerbugs
- bodhi
- bugzilla2fedmsg
- busgateway
- datagrepper
- dbserver
- dns
- fedimg
- github2fedmsg
- ipa
- kernel-qa
- kerneltest
- kojibuilder
- kojihub
- kojipkgs
- logging
- mailman
- memcached
- mirrormanager
- nagios
- notifs
- oci-registry
- odcs
- openqa
- openqa-workers
- osbs
- packages
- pdc-web
- pkgs
- proxies
- rabbitmq
- releng-compose
- resultsdb
- secondary
- sign-bridge
- sundries
- value
- wiki

Signed-off-by: Nils Philippsen <nils@redhat.com>
2021-02-01 22:23:41 +00:00

909 B

— nm: 255.255.255.0 gw: 10.3.166.254 dns: 10.3.163.33 ks_url: http://10.3.163.35/repo/rhel/ks/kvm-rhel-8-iad2 ks_repo: http://10.3.163.35/repo/rhel/RHEL8-x86_64/ volgroup: /dev/vg_guests eth0_ip: 10.3.166.59 vmhost: vmhost-x86-10.stg.iad2.fedoraproject.org datacenter: iad2

databases:

  • datanommer

dbs_to_backup:

  • datanommer

lvm_size: 1000000 mem_size: 65536 max_mem_size: 98304 num_cpus: 8 tcp_ports: [ 5432, 443 ]

kernel_shmmax: 68719476736

db_backup_dir: ['/backups'] shared_buffers: "4GB" effective_cache_size: "12GB"

sar_script: /usr/local/bin/datagrepper_sar.py sar_script_user: root sar_output_file: datagrepper.csv sar_huge: true