No. Que does not balance jobs across satellites. Our FAQ says Que does not offer load balancing of jobs at this time. You assign jobs to specific satellites yourself, and before you do, you check that every font, plugin and asset the job needs is on the machine that will render it.
That sounds like a gap, and for some pipelines it is. For others it is a perfectly workable arrangement, as long as you plan the split instead of expecting the system to plan it for you. This page explains how assignment works, how to split a batch across machines, and what to check before you trust a second render machine.
How a job ends up on a satellite
Start with the vocabulary. A satellite is a machine that has been added to your Que account so it can process jobs. Templater has to run on that machine. When Templater adds the satellite, it profiles the host machine and sends the profile to Que. A job is one set of properties, stored in a campaign, that describes one versioning and render task. A campaign is the container for a group of jobs. If these terms are new, what a Que satellite is and how to add one covers the setup.
Every job carries a key named bot, and its value is the ID of the satellite assigned to it. That key is the whole assignment mechanism. There is no scheduler sitting above it that looks at how busy each machine is and hands out work. The satellite named in the job is the satellite that will process it.
You do not have to set that key on every job by hand, because campaigns have defaults. When you create a campaign, you choose a default autograf and a default satellite. The API documentation says that when you create a job without an aep or bot value, Que sets the job to the campaign’s default autograf and satellite. The batch import behaves the same way. If an imported job object has no aep or bot key, Que adds them using the campaign’s defaults. So the simplest setup is one campaign, one default satellite, and every job lands on that one machine.
How jobs render
Assignment comes before rendering starts
-
Add the satelliteTemplater adds the satellite, profiles the host machine, and sends the profile to Que.
-
Set the bot valueEvery job carries a key named bot, and its value is the ID of the satellite assigned to it.
-
Use campaign defaultsWhen you create a job without an aep or bot value, Que sets the job to the campaign’s default autograf and satellite.
-
Authorize the satelliteYou assign jobs to an authorized satellite, then you enable Templater’s Bot functionality on that machine.
-
Bot processes ready jobsThe Bot begins processing the jobs that have a render-status of ready.
Once more than one satellite exists, you have three ways to direct jobs. The first is to choose the satellite for a set of jobs in the Que Dash. Our help pages say that when you have satellites in your account you can assign them to process the jobs you select. The second is to set the bot value in each job when you create it or import it through the API. The third is to update an existing job with the API. The update endpoint takes the keys you want to change and their new values, so changing bot to a different satellite ID should reassign the job. This is our reading of the API docs, which show the update endpoint with a different key (output) and never show a bot change, so test it on a single job first.
What happens after assignment
Assignment alone does not render anything. The API documentation describes the order this way. You assign jobs to an authorized satellite, then you enable Templater’s Bot functionality on that machine, and the Bot begins processing the jobs that have a render-status of ready. The machine only handles jobs that name it, and Templater’s job selection can filter further by job ID, render status, properties or creation date. There is also a processing scope on each satellite. It says whether the satellite processes jobs from all campaigns or only from the single campaign linked to Templater. That setting matters when two campaigns share a machine, and it matches a checkbox in Templater’s Que configuration dialog.
You can also see the assignment from the satellite side. The API has an endpoint that returns a paginated list of all jobs assigned to a satellite, and it requires that satellite’s own access token. Campaign statistics, in the Que Dash and through the API, show which satellites are assigned to jobs in a campaign and how many jobs each satellite has processed. Those numbers are how you tell, after the fact, whether your split did what you intended.
Check the machine before you assign to it
The FAQ answer has a second sentence that carries as much weight as the first. You must verify that all required assets and resources, including fonts and plugins, are installed or available on any satellite you assign a job to. Our help page on working with satellites adds the underlying requirement. To render a job, a satellite must have the job’s referenced autograf installed, along with any dependencies such as fonts and effects.
The reason is simple. A job is data. It does not carry the After Effects project, the typefaces or the effects with it. The render machine opens the project locally and draws the data into it. If the typeface that the title layer uses is not installed on that machine, the machine cannot draw the title the way your authoring workstation does. So adding a satellite to the account is a small step, and making that machine equal to the first one is the real work.
The Que Dash gives you a way to compare. Expand a satellite and you can read the profile Templater sent. The host properties include the operating system, installed RAM, the CPU type, the drive capacity and a list of all installed fonts. The engine properties include the authoring app and its version, the Templater version, the processing scope, the output mode, the output naming scheme, the output root, the footage root and a list of installed effects. That is a ready-made checklist. Put two satellites side by side and look for differences in fonts, effects, the After Effects version and the Templater version before you split a batch between them.
Satellite profile
What the Que Dash shows for each satellite
| Property | Host profile | Engine profile |
|---|---|---|
| Operating system | Listed under host properties | — |
| Installed RAM & CPU | RAM, CPU type, drive capacity | — |
| Installed fonts | Full font list | — |
| Authoring app & version | — | App and its version |
| Templater version | — | Templater version |
| Output & footage roots | — | Output root and footage root |
Two of those properties deserve extra attention because they involve paths. The output root is where the satellite saves finished files, and the footage root is where it looks for footage assets. If one machine saves to a local drive and another saves to a shared volume, your finished files will end up in two places. If footage lives at one path on the first machine and a different path on the second, jobs that reference footage may fail on the second. Making the roots match, or deliberately choosing where each machine writes, is part of the setup. Our help pages also suggest running a sync folder service on each satellite, so the autograf stays current between the authoring workstation and every render machine. When you change the template, every satellite needs the new version, or half your outputs will come from the old design.
Before you assign
Verify a second satellite before sending jobs
- Confirm the job’s autograf is installed on the satellite
- Verify all required fonts are installed on the machine
- Verify all required effects and plugins are installed
- Match the After Effects version to your first machine
- Match the Templater version to your first machine
- Check that the footage root and output root paths match
Slots and plan limits
You also have a ceiling on how many machines can take part. The plan table in the Que help pages lists two satellites on the Solo plan and five on the Team plan. Our help pages say you can purchase more slots if you need them for a larger fleet. If you try to add a satellite and the Add to Que button fails, one documented cause is that you have reached the satellite limit for your plan. Another is a name conflict. Each satellite must be uniquely named, and its name comes from the host machine, so a rename means renaming the host and then syncing the machine profile. The troubleshooting steps are in our article on troubleshooting Que satellites.
Plan satellite limits
Solo allows two satellites, Team allows five
- Satellites included
- Solo planTwo satellites
- Team planFive satellites
- Need more machines?
- Solo planPurchase additional slots
- Team planPurchase additional slots
You can purchase more slots if you need them for a larger fleet.
Splitting a batch by hand, with an example
Because the split is yours, design it. The following is general practice, not a documented feature, and the numbers in the example are invented.
Suppose you have a campaign of 600 jobs that will render on two satellites, one called STUDIO-A and one called STUDIO-B. You have checked that both machines have the same version of After Effects and Templater, the same fonts and effects, the same footage root, and the same output root pointing at one shared location. Now you need to decide which jobs go where.
The simplest rule is a plain alternation. Job one goes to STUDIO-A, job two to STUDIO-B, and so on. You can do this when you import, because the import accepts a bot key on each job object. Your script writes the satellite ID into the bot field as it builds the list. It is a few lines of code, and it spreads the count evenly.
A plain alternation spreads the number of jobs, not the amount of work. If some jobs are long and some are short, a count-based split can leave one machine busy long after the other has finished. When your jobs differ in length, group them so that each machine gets a similar mix, or hold some jobs back and assign them later to whichever machine finishes first. That second approach is manual load balancing, and it is the work the platform does not do for you. Que’s job states make it practical, because you can watch the ready, queued, processing and done counts per campaign and decide what to hand to another machine. Our article on Que job states and what to do next explains what each status means.
Another approach is to split by campaign instead of by job. Make one campaign per machine, set each campaign’s default satellite to its machine, and import each group into its own campaign. The assignment is then implicit and you never write a bot value. The cost is that a single logical project now lives in several campaigns, so your reporting has to add them back together. Pick whichever is easier for your team to reason about at three in the morning when a batch is running late.
Then run a canary first. Send a few jobs to the new machine alone, render them, and look at the output next to the same jobs rendered on the first machine. You are looking for fonts that fell back, effects that did not apply, and files that landed in the wrong folder. This costs a few minutes and saves the discovery of a bad batch after the fact. Our article on designing a canary row goes through the idea in more detail.
Who this arrangement suits, and who it does not
Manual assignment suits a team that knows its own machines. If you have two or three render computers that you configured yourselves, you can keep them identical, you can predict which jobs run long, and a short script or a few clicks in the Que Dash handles the split. You also gain control.
It suits a team with an unpredictable stream of arrivals much less well. If jobs arrive all day from an application and nobody is watching, a fixed assignment means the work queues behind whichever machine you picked, even when another machine sits idle. In that situation you would have to write your own dispatcher on top of the API, choosing a satellite for each new job based on what you can read from the campaign statistics. That is possible, because the pieces are in the API. It is also code you own and maintain, so weigh it honestly against your needs.
What to do when a satellite is removed
One more documented behavior affects how you plan. When a satellite is deleted, every job assigned to it becomes unassigned, and any campaign that used it as the default must be changed to a different default satellite. Jobs that are unassigned will not process until you assign them to another authorized satellite. This page does not cover what happens to a job if a machine goes offline partway through a render. Ask our support team about that case before you build a plan that depends on a particular answer.
Bottom line
Que gives you assignment, not balancing. If one machine can keep up with your batches, set one default satellite and move on. If you need two or more, treat the split as part of your pipeline design. Match the machines, assign deliberately, watch the per-satellite counts, and rebalance by hand. If you want a system that distributes work across a pool without your involvement, tell us what you are trying to do before you buy, because that is not what Que does today. For the larger picture of where Que sits, read how Que and Templater work together.
If you are weighing a multi-machine setup, these are the questions to settle before you plan the split.
Frequently Asked Questions
Does Que load balance jobs between satellites?
No. Our FAQ says Que does not offer load balancing of jobs at this time. Each job names the satellite that will process it, either through the campaign default or through a bot value you set. You decide how a batch is divided between machines. If you need automatic distribution across a pool, raise that with us before you commit to a pipeline.
How do I assign a job to a specific satellite?
You have three routes. Select jobs in the Que Dash and assign a satellite to them, set the bot key to the satellite ID when you create or import a job through the API, or update an existing job with the API and change its bot value. The API docs do not show that last call with bot, so test it on a single job first. If you set nothing, Que uses the default satellite chosen for the campaign.
What should I check on a second satellite before assigning jobs?
Confirm it has the job’s autograf installed along with its dependencies, including fonts and effects. Compare the profile in the Que Dash against your first machine, looking at fonts, effects, After Effects version, Templater version, footage root and output root. Then send a few test jobs and inspect the finished files.
How many satellites can I have?
The Que help pages list two satellites on the Solo plan and five on the Team plan, and say you can purchase additional slots. Each satellite needs a unique name, which comes from its host machine. If adding a satellite fails, the limit and a name conflict are both documented causes. See the Que page or ask our team for current details.
Can I see which satellite processed which jobs?
Yes. The Que Dash campaign statistics show which satellites are assigned to jobs in a campaign, how many jobs each has processed and when its last output was created. The API can also list the jobs assigned to a satellite, using that satellite’s own access token. Use those numbers to check that your split matches your plan.
What happens to jobs if I delete a satellite?
They become unassigned, and any campaign that used that satellite as its default needs a new default. Unassigned jobs will not process until you assign them to another authorized satellite. Our pages do not describe what happens if a machine goes offline during a render, so ask support about that scenario.
