Related to swh/infra/sysadm-environment#4739

Refs. swh/infra/sysadm-environment#4839

So it's always ready to ingest fast a new demand.

Refs. swh/infra/sysadm-environment#4834

Refs. swh/infra/sysadm-environment#4834

Revert "staging: Stop loader metadata"

This reverts commit 133ecfb3.

According to what we agreed to configure in production.

Refs. swh/infra/sysadm-environment#4830

Refs. swh/infra/sysadm-environment#4830

Related to swh/infra/sysadm-environment#4739

Refs. swh/infra/sysadm-environment#4685

- Decrease the directory replayers because the replay is almost done
- Increase the content replayers to speed up the replay

Related to swh/infra/sysadm-environment#4739

Refs. swh/meta#4949

Which will allow to dispatch correctly issues to sentry even for applications with
multiple loaders within (save-code-now/highpriority, add-forge-now).

Refs. swh/meta#4949

Refs. swh/meta#4949

Refs. swh/meta#4949

Refs. swh/meta#4949

Refs. swh/meta#4949

Refs. swh/meta#4949

Refs. swh/meta#4949

This reverts commit f36ae837.

This aligns how the configuration secrets are installed inside the main configuration
file (using awk now).

Refs. swh/meta#4949

Refs. swh/meta#4949

Related to swh/infra/sysadm-environment#4739

Their lag slowly but steadily slightly increases [1]

[1] https://grafana.softwareheritage.org/goto/UkYHiQBVk?orgId=1

Refs. swh/infra/sysadm-environment#4794

Its lag currently increases [1]

[1] https://grafana.softwareheritage.org/goto/Ni3a1EB4z?orgId=1

Refs. swh/infra/sysadm-environment#4758

Related to swh/infra/sysadm-environment#4822

As the version has a '.' inside, helm transforms it to a float

```
couldn't parse image reference "container-registry.softwareheritage.org/swh/infra/swh-apps/graphql:2.02303143e+07": invalid reference format
```

Refs. swh/infra/sysadm-environment#4820

Refs. swh/infra/sysadm-environment#4820

Refs. swh/infra/sysadm-environment#4820

Refs. swh/infra/sysadm-environment#4794

This reverts commit 510b86a9.

Refs. swh/infra/sysadm-environment#4809

Since we added the redis configuration, we need to reset topics so we get back the
original errors.

Refs. swh/infra/sysadm-environment#4809

Refs. swh/infra/sysadm-environment#4809

regarding unprivileged topics and redis error reporter use.

Refs. swh/infra/sysadm-environment#4809
Refs. swh/devel/swh-storage#4676

Refs. swh/infra/sysadm-environment#4809

Refs. swh/infra/sysadm-environment#4813

It was a test to ensure the autoscaling worked.

Refs. swh/infra/sysadm-environment#4794

It will allow to recover the cassandra monitoring service
which lost the endpoints from time to time (it's the same for the
production)

Related to swh/infra/sysadm-environment#4809

No need to pass by public host

Refs. swh/infra/sysadm-environment#4794