Commits · 290820410c8c1f40038d5537436bcae3745b228d · Platform / Infrastructure / sysadm-provisioning · GitLab

Snippets Groups Projects

Aug 25, 2023
- admin/cluster-admin.tf: Add resource dependencies before cluster removal · 29082041
  Vincent Sellier authored 1 year ago
  
  Related to sysadm-environment#4730
  Verified
  
  29082041
- admin/cluster-admin-rke2: Fix thanos certificate generation · fe443b13
  Vincent Sellier authored 1 year ago
  
  Related to sysadm-environment#4730
  Verified
  
  fe443b13
- staging: add rancher-monitoring dependencies on test cluster · 132ddea0
  Guillaume Samson authored 1 year ago
  
  Related to sysadm-environment#5018
  132ddea0
Aug 24, 2023
- admin: Create the admin-rke2 cluster to replace the admin rke1 one · 706aca29
  Vincent Sellier authored 1 year ago
  
  Related to sysadm-environment#4730
  Verified
  
  706aca29
Aug 23, 2023

staging: Add the test cluster definition · 3a2549b5

Vincent Sellier authored 1 year ago

This cluster was used to test the kubernetes upgrade before doing
it for real in the archive-staging-cluster.

Such cluster can be useful to test some workload or migrations
more easily than in a local cluster. So let's keep it for now.

The configuration allow the full creation in one plan/apply command
without human actions. It's possible with the definition of the
vmid parameter for the vms and the creation of the snapshotter configuration
file in the post provision step

Related to sysadm-environment#5014

3a2549b5

Aug 22, 2023

production/rancher: upgrade Kubernetes to the last version · f28b58e7

Guillaume Samson authored 1 year ago

From kubernetes_version 'v1.24.9+rke2r2' to 'v1.24.16+rke2r1' then to 'v1.25.12+rke2r1' then to 'v1.26.7+rke2r1'
Related to sysadm-environment#4999

f28b58e7

production/archive-production-rke2: disable worker nodes drain on upgrades · 1b40c146
Guillaume Samson authored 1 year ago
```
Related to sysadm-environment#4999
```
1b40c146

staging/archive-staging-rke2: Align terraform with the cluster configuration · 9c4635df

Vincent Sellier authored 1 year ago

Some minor changes after the migration to 1.26.
Ensure the default rancher ingress controller is not deployed

Related to sysadm-environment#4999

9c4635df

staging/archive-staging-rke2: Disable node drain on upgrades · 19fa537f

Vincent Sellier authored 1 year ago

It seems it's not always necessary and could create a lot of
downtime.
We'll need to check depending of the kind of upgrade we apply.

Related to sysadm-environment#5014

19fa537f

production/rancher: upgrade cattle-monitoring-system · 151988f7

Guillaume Samson authored 1 year ago

From '100.1.3+up19.0.3' to '101.0.0+up19.0.3' then to '102.0.1+up40.1.2'.
Related to sysadm-environment#4999

151988f7

Aug 21, 2023
- staging/rancher: Upgrade the kubernetes version to the last available 1.24 · 6f8e2485
  Vincent Sellier authored 1 year ago
  
  Related to sysadm-environment#5014
  Verified
  
  6f8e2485
Aug 18, 2023
- staging/kube: Upgrade monitoring app to last one available · 17b7e871
  Vincent Sellier authored 1 year ago
  
  it removes the usage of PosSecurityPolicies object in the cluster Related to sysadm-environment#5014
  Verified
  
  17b7e871
- admin/kubernetes: align the kubernetes version with the reality · f6a4dd1c
  Vincent Sellier authored 1 year ago
  
  Related to sysadm-environment#4999
  Verified
  
  f6a4dd1c
Aug 17, 2023
- admin: upgrade prometheus application to last version · fdffbd62
  Vincent Sellier authored 1 year ago
  
  It's necessary to use a 1.25+ compatible version Related to sysadm-environment#4999
  Verified
  
  fdffbd62
- admin: Upgrade admin cluster version to match the reality · c526c653
  Vincent Sellier authored 1 year ago
  
  Related to sysadm-environment#4999
  Verified
  
  c526c653
- admin: Upgrade admin cluster version to match the reality · 9b2eeb61
  Vincent Sellier authored 1 year ago
  
  Related to sysadm-environment#4999
  Verified
  
  9b2eeb61
Jul 27, 2023
- staging/archive-rke2-staging: add worker6 node · 1336aa4c
  Guillaume Samson authored 1 year ago
  
  1336aa4c
Jul 10, 2023
- cluster-staging-rke2: Increase limits of the prometheus pod · 7a8964e9
  Vincent Sellier authored 1 year ago
  
  Closes sysadm-environment#4980
  Verified
  
  7a8964e9
May 30, 2023
- production: Increase webapp1 memory after several oom killer · 38461b92
  Vincent Sellier authored 1 year ago
  
  Verified
  
  38461b92
May 19, 2023
- staging: Add new worker node · 8127c025
  Antoine R. Dumont authored 1 year ago
  
  Refs. sysadm-environment#4883
  Verified
  
  8127c025
- staging: Bump mgmt node memory · 7770441f
  Antoine R. Dumont authored 1 year ago
  
  Refs. sysadm-environment#4883
  Verified
  
  7770441f
Apr 24, 2023
- Remove useless retention duration on the rancher clusters · 04ed4284
  Vincent Sellier authored 1 year ago
  
  The indentation was wrong so the configuration was not applied
  Verified
  
  04ed4284
- production/archive-production-rke2: Increase prometheus resources · 1ba4988a
  Vincent Sellier authored 1 year ago
  
  Given the number of resources on the cluster, the prometheus container and the node exporters are regularly restarted due to resource overheads sysadm-environment#4866
  Verified
  
  1ba4988a
- production/archive-production-rke2: sync kubernetes version according to the... · 40fff580
  Vincent Sellier authored 1 year ago
  
  production/archive-production-rke2: sync kubernetes version according to the current deployed version
  Verified
  
  40fff580
Apr 06, 2023
- Decommission gsa-temp staging node · 1460220a
  Guillaume Samson authored 1 year ago
  
  1460220a
Mar 24, 2023
- Decommission worker0 staging node · 0068053d
  Antoine R. Dumont authored 1 year ago
  
  Refs. sysadm-environment#4818
  Verified
  
  0068053d
Mar 23, 2023
- staging/workers: bump scratch torage to 100G · 2d3965ab
  Guillaume Samson authored 1 year ago
  
  Related to swh/infra/puppet/puppet-swh-site#3565
  2d3965ab
Mar 22, 2023
- admin/dali: Bump storage to 350g · 8801a502
  Antoine R. Dumont authored 1 year ago
  
  Refs. sysadm-environment#4816
  Verified
  
  8801a502
Mar 08, 2023

archive-staging-rke2: fix prometheus ingress configuration · d5dbe729

Vincent Sellier authored 2 years ago

- The vhost was not matching what is declared in the thanos configuration
- Use the let's encrypt production issuer (a valid certificate is needed by thanos)

Related to sysadm-environment#4797

d5dbe729

Mar 03, 2023
- update gsa-temp-vm cpu in staging env · 34f89938
  Guillaume Samson authored 2 years ago
  
  needed to start vagrant vm related to sysadm-environment#4790
  34f89938
- update kubernetes version staging environment · 07dac6bc
  GuiSam authored 2 years ago and Guillaume Samson committed 2 years ago
  
  07dac6bc
- create gsa-temp-vm in staging env · 8d799322
  GuiSam authored 2 years ago and Guillaume Samson committed 2 years ago
  
  to be used until I receive my laptop related to sysadm-environment#4790
  8d799322
Mar 02, 2023
- doc: add the rancher token configuration in the setup script · 4e051716
  Vincent Sellier authored 2 years ago
  
  Verified
  
  4e051716
Jan 27, 2023
- staging-rke2: Increase worker nodes' cores to 6 · 9251aa85
  Antoine R. Dumont authored 2 years ago
  
  Verified
  
  9251aa85
- staging: Make nodes restart on boot · 98099d39
  Antoine R. Dumont authored 2 years ago
  
  Now that the intervention is done.
  Verified
  
  98099d39
Jan 24, 2023

Make staging nodes not restart on boot · d4399e77

Antoine R. Dumont authored 2 years ago

This is to prepare the hardware intervention starting 25/01/2023.

This also aligns the current staging nodes with their actual hypervisors.

Refs. sysadm-environment#4742

d4399e77

Jan 10, 2023
- admin: Increase cpu limits for prometheus to stop the CPUThrottlingHigh flapping alert · 1d8fd55c
  Vincent Sellier authored 2 years ago
  
  Related to CPUThrottlingHigh
  Verified
  
  1d8fd55c
- archive-staging: Decomission cluster · 1225ce61
  Antoine R. Dumont authored 2 years ago
  
  Ref. sysadm-environment#4699
  Verified
  
  1225ce61
- archive-staging: Decomission worker nodes · 7b2fa4e9
  Antoine R. Dumont authored 2 years ago
  
  It got done progressively, from worker4 to 1 (`terraform destroy -target=...`). And then from pergamon, swh-puppet-master-decomission then `puppet agent --test` (to drop icinga alerts). Ref. sysadm-environment#4699
  Verified
  
  7b2fa4e9
Jan 09, 2023
- archive-staging-rke2: Add a new worker · 0b58566f
  Antoine R. Dumont authored 2 years ago
  
  Ref. sysadm-environment#4699
  Verified
  
  0b58566f