Running the portal¶
For the people who install and run a Claugine portal: how it is built, how it gets new releases, and how it is backed up and restored.
Standard and Enterprise¶
The two editions run the same portal; they differ in how much of it can be changed. Where the pages below differ by edition, they say so:
| Standard | Enterprise | |
|---|---|---|
| Portal nodes | one app VM | two, active and standby, on one or two sites |
| Backups | local, to ~/claugine/backups, once a day at 03:06, kept 3 days; S3 off |
the folder, the time and how long they are kept, and backups to S3 |
| Updates | stable releases only, in one window a week: Sunday, 07:02 | every release stream, the time chosen, versions held |
| Editable backup settings | no | yes |
| Editable update settings | no | yes |
| Preview release stream | no | yes |
| Code | public, fetched by the node itself | private, delivered by my.claugine.com |
In the current version, Standard pins these eight settings, and Settings shows them locked:
| Setting | Pinned value |
|---|---|
| Local backups | on |
| Backup folder | ~/claugine/backups (~ is the account that ran the installation) |
| Backup time | 03:06, once a day |
| Backups kept | 3 days |
| S3 backups | off |
| Release stream | stable |
| Update day | Sunday |
| Update time | 07:02 |
They are applied when the portal starts and again after a restore, so editing the database by hand does not survive. Everything about a second node (the standby, replication, the HA agent, the load balancer, switching over) applies to Enterprise only.
| Page | What it covers |
|---|---|
| Portal architecture | what a Standard node runs; the Enterprise design: the app VMs and their roles (active, standby), the database and its replication, HAProxy on each site, the application and its timers, the HA agent, secrets, mail and certificates |
| Software delivery and updates | where a node gets the portal; Standard's weekly update window; Enterprise's two nodes: the rules, automatic and by hand in the portal |
| Backup and restore | what a backup holds, when and where it is made, the archive password, the platforms' front-end backups, a full restore and its checklist |
| Installation | installing the portal with the installation script, and what to do after it |
flowchart LR
I[Install with the script] --> R[Run]
R --> U[Update: Standard weekly; Enterprise standby first]
U --> R
R --> B[Back up: daily]
B -. "lost a VM" .-> RS[Restore on a new VM]
RS --> Rflowchart LR
I[Install with the script] --> R[Run]
R --> U[Update: Standard weekly; Enterprise standby first]
U --> R
R --> B[Back up: daily]
B -. "lost a VM" .-> RS[Restore on a new VM]
RS --> R
What the portal does for the provider's staff and customers is in Portal and the Portal reference.