AdminBolt 1.9.2: Background Jobs, Servers With Many Sites and PHP Settings

AdminBolt 1.9.2 is out. A new Background Jobs page shows what the panel runs in the background. Failed jobs can be run again. Servers with many sites get more room to grow. Every setting on a PHP version's configuration page now reaches the server.

Background Jobs page

Health > Background Jobs lists the jobs that are waiting or running and the jobs that failed, in plain words. For a failed job you see what went wrong and when. The technical details sit behind Details, with sensitive values masked.

You can run a failed job again or delete it. Run all again runs the failed jobs again, a batch at a time. Delete broken jobs clears out the jobs that can never run again. The admin REST API offers the same actions under /api/background-jobs. When jobs fail or pile up, the dashboard and Health point you to the page. Long tasks, such as the package check after a system package upgrade, run to completion.

Servers with many sites

Apache, LiteSpeed and PHP-FPM keep serving reliably as you add more sites to a server. Existing servers get this when they upgrade. The services do not restart for it.

AutoSSL handles servers with many domains. The AutoSSL tab shows how far it has got: domains done, issued, failed, skipped and left for a later pass. On the hub, the DNS sync stays quick on hubs with thousands of zones.

PHP settings that reach the server

Every php.ini and extension setting on a PHP version's configuration page reaches the server. That includes Implicit Flush, the OPcache settings and the Imagick settings. The page reports the configuration as applied once the server uses it.

Remi PHP, alt-php and LSPHP of the same release are now separate versions. Each one lists only the websites that run it. For a domain, the PHP-FPM Configuration tab of PHP Domain Assignment shows the environment variables, the ping settings and the terminate timeout. What you save there takes effect. Only administrators can set these values. Every alt-php extension switch reaches the server and saved alt-php settings take effect at once (CloudLinux Only).

A failed LiteSpeed Enterprise, OpenLiteSpeed or LSAPI Engine install names the missing package and the next step. Installing one of them sets up LSPHP for every installed Remi PHP version (AlmaLinux Only). Uninstalling LiteSpeed Enterprise sets what it left behind aside instead of deleting it, so a later OpenLiteSpeed install starts clean. The Web Server pages show only the tabs that apply to your web server.

Secondary DNS on the hub

Each secondary DNS server in a cluster receives new zones. SRV records resolve on the hub and on the secondary DNS servers. When the Secondary DNS role cannot be turned on for a node, the panel shows the node's reason. Register again repairs a node's registration.

Smaller things you may notice

SSH Access and Resolver Configuration show the server's current settings and save only the fields you changed. Turning root login off asks for confirmation. Changing the SSH port also opens the new port in the firewall.

The WordPress installer says so when the plan has no free database left, before it changes anything. Go to Admin signs in to WordPress through single sign-on, also while you impersonate the account. The autoresponder schedule names its time zone. The admin phpMyAdmin settings page is gone. Administrators open the panel databases with Adminer and customers keep phpMyAdmin.

How to update

Upgrades run from the panel. Open System, then System Updates, check for updates and start 1.9.2 from there. New installs are one command:

curl -sSL https://get.adminbolt.com/install.sh | bash

Tell us how it goes

Tell us how 1.9.2 works for you on Discord or the community forum.