Platform status
Check at a glance whether Elion ITSM's services, databases and job queues are working, and what to look at when something falls behind.
On this page
The Platform status page shows the health of Elion ITSM as a whole, not of one company: whether its services respond, how much space its databases use and how the jobs running in the background are progressing, like sending emails or SLA notifications.
Before you start
You need to be a platform superadministrator. Read Platform administration.
Open the page
Enter the administration with Platform administration, at the bottom of the side menu.
In the side menu, choose Platform status.
The page refreshes on its own every 30 seconds while you have it in view. To refresh it right away, click Refresh.
The indicators at the top
Service indicators show OK in green or Down in red. The Companies indicator shows the total number of companies.
Indicator | What it shows |
|---|---|
api | The service that serves the web application and the API, with how many instances are running (inst.) and its version. |
worker | The service that processes background jobs (the queues below). |
ws | The real-time service, which updates screens without reloading. |
Redis | The in-memory database that stores, among other things, the job queues. It shows the memory it uses in MB. |
PostgreSQL | The main database, with its size in MB and the open connections (conn.). |
Companies | How many companies there are. If any has an outdated data structure, it turns red and says how many have a failed migration. |
A service shows Down when none of its instances has reported recently that it is alive. Hover over the indicator to see the details, for example No recent heartbeat.
The queue table
Each row is a job queue. The columns are:
Column | What it counts |
|---|---|
Queue | The queue name. |
Waiting | Jobs ready that have not started yet. |
Active | Jobs being processed right now. |
Delayed | Jobs scheduled for later and retries waiting their turn. |
Failed | Jobs that failed after using up their retries. It is shown in red if it is greater than 0. Each failure stops being counted after one day. |
Completed (1 h) | Jobs that finished successfully in the last hour. |
What each queue does
Queue | What it processes |
|---|---|
scheduler | Periodic tasks: SLA checks, automatic ticket closing, knowledge page reviews, holidays and daily sprint snapshots, among others. |
Sending emails: ticket notifications and links to create or reset passwords. | |
notify | Notifications by email, browser notifications, webhooks and Telegram, and escalations. |
sla | Alerts for SLAs about to be breached and breached, and target recalculations. |
workflow | Running BPMN workflows: starts, completed tasks, timers and automatic actions. |
mailin | Emails that reach the desks and are turned into tickets or replies. |
ai | The assistant's background tasks: classifying new tickets, summaries, article drafts and search indexes. |
hooks | Argunis Monitoring webhooks, sending evidence to SofIA Governance and checking help center custom domains. |
agile | Daily sprint snapshots, released version notifications and agile desk metrics. |
projects | Recalculating project schedules. |
maintenance | Maintenance: cleaning up old data, rotating encryption keys, importing users, reindexing search, monthly customer reports and contract renewals, among others. |
What to check if something falls behind
What you see | What it means | What to do |
|---|---|---|
worker shows Down | No service is processing the queues. The Waiting numbers will grow in all of them. | It is the first thing to bring back up. Let the Argunis technical team know immediately. |
Redis shows Down | Without Redis, queues are not processed. The queue table may appear empty with No queues. | Let the technical team know immediately. |
Waiting grows and Active stays at 0 | The worker is not taking jobs from that queue. | Check that worker is OK. If it is, let the technical team know with the queue name. |
Waiting is high, but Completed (1 h) also goes up | There is a lot of work, but it is moving. This is normal after a large import or a wave of emails. | Wait a few minutes and click Refresh. |
Delayed grows in mail or notify | Sends are failing and being retried with longer and longer gaps. | Check whether there is a problem with the email provider or with the webhook destinations. If it continues, let the technical team know. |
Failed greater than 0 | Jobs that could not be completed even with retries. For example, emails that did not go out. | Note the queue and the approximate time and let the technical team know so they can review the details. |
Companies in red with failed migration | One or more companies have an outdated data structure. | Look in Company list for the ones with the migration tag and let the technical team know. |
If a customer reports that notifications are not reaching them, look at the mail and notify queues first. If a customer says their emails to the desk do not create tickets, look at mailin.
If the page shows Couldn't load: followed by a reason, the api service might have problems: try again in a moment and, if it continues, let the technical team know.