> For the complete documentation index, see [llms.txt](https://docs.terrakube.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.terrakube.io/2.34.0/getting-started/deployment/executor-log-output.md). # Executor Log Output What the executor does with the output of a step (the lines Terraform, OpenTofu and scripts print), and the settings that control it. The defaults suit most installations. {% hint style="info" %} Available from Terrakube 2.34.0. {% endhint %} ### Where output goes * **Saved output**: every line, whole, written to storage when the step ends. This is what the **Console** tab and the download show after the step. * **Live stream**: what viewers read while the step runs, through Redis (or the API when `io.executor.log-via-api=true`). It is bounded and may fall behind, but never slows the run. * **Executor service log** (`kubectl logs`): **off by default since 2.34.0**. Output was previously copied here too, which doubled log ingestion cost and exposed printed values to anyone who can read cluster logs. Set `ExecutorLogsEchoToServiceLog=true` to restore it. ```mermaid flowchart LR X[Executor step output] -->|live, bounded| R[(Redis stream)] R --> UI[Console while running] X -->|parts every 1 MiB or 30 s| ST[(Object storage)] X -->|final object when step ends| ST ST --> UI2[Console and download after the step] X -. API outage: final snapshot spooled .-> ST ``` ### Output survives executor failure While a step runs, the executor also copies its output to storage in numbered parts (`tfoutput///.parts/`). They are deleted when the final output is saved. If the executor dies first, the API serves the parts joined in order and the Console shows a warning that the output is incomplete. If the API is unreachable at the end of a plan or apply, the final structured result is spooled to storage and applied once the API returns, without re-running Terraform. ### Settings Environment variables on the executor unless noted. | Variable | Default | Meaning | | ----------------------------------------- | ---------- | -------------------------------------------------------------------------------------------------------------------------------------------- | | `ExecutorLogsEchoToServiceLog` | `false` | Also write output to the executor's own log. | | `ExecutorLogsLiveLineMaxBytes` | `65536` | Longest line sent to the live stream; longer lines are cut there only. | | `ExecutorLogsStepOutputMemoryBytes` | `4194304` | Output kept in memory per step before moving to a temporary file. A step printing more than this needs scratch space. | | `ExecutorLogsRenderedPlanContextMaxBytes` | `49152` | How much of a plan's text is published for the pull request comment. A longer plan is cut with a note; the step output keeps the whole plan. | | `ExecutorLogsArchiveEnabled` | `true` | Copy output to storage in parts while the step runs. | | `ExecutorLogsArchivePartBytes` | `1048576` | Write a part once this many new bytes exist. | | `ExecutorLogsArchiveIntervalSeconds` | `30` | Also write a part after this long if there is anything new. | | `ExecutorLogsArchiveMaxParts` | `1000` | Most parts per step. | | `ExecutorLogsRedisMaxLen` | `200000` | Entries kept in a job's live stream. | | `ExecutorLogsApiQueueMaxBytes` | `16777216` | With `log-via-api`, bytes queued while the API is slow; oldest lines are dropped from the live view first. | | `ExecutorStructuredOutputQueueMaxBytes` | `67108864` | Memory budget for structured-output snapshots waiting to be saved. | | `LogsAppendMaxBytes` (API) | `4194304` | Largest log request the API accepts. | Every limit affects only the live view. The run and the saved output are never trimmed. ### Run limits Runs that stay open too long are closed by Terrakube. Set these on the API: | Variable | Default | Applies to | | ----------------------------------------- | --------- | -------------------------------------------------------------------------------------------------------------------------------------- | | `InactiveJobsMaxAgeHours` | `6` | Pending, approved, queued and running runs are closed as failed. Time spent waiting for approval or for an earlier run does not count. | | `InactiveJobsWaitingApprovalTimeoutHours` | `0` (off) | Runs waiting for approval are cancelled after this many hours. | --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation by asking a question. Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters: ``` GET https://docs.terrakube.io/2.34.0/getting-started/deployment/executor-log-output.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.