> ## Documentation Index > Fetch the complete documentation index at: https://www.octoparse.com/docs/llms.txt > Use this file to discover all available pages before exploring further. # Waits, timeouts, and retries > Use Octoparse waits, timeouts, and action recovery to make a task more reliable on slow or asynchronous pages. An Octoparse task moves through a website one browser action at a time. It opens a page, clicks an element, scrolls, or enters text; each action changes the page state that the next action sees. Waits, timeouts, and retries control when the task moves on, when it stops waiting, and when it tries a supported action again. They help with pages that load data after a click, scroll, or background request. They do not fix an invalid selector, an expired login session, or a website that refuses access. ## How browser actions move a task forward | Browser action | What happens in the browser | When to use extra control | | - | - | - | | **Go to Web Page** | Opens a URL and waits for the page to load | The page can redirect, load slowly, or continue loading data after the first page load | | **Click Item** or **Click to Paginate** | Clicks an element to open a detail page, reveal content, or move to the next page | The click can trigger a delayed page change or background request | | **Scroll** | Moves through a page to expose more content | Infinite-scroll pages can add items after each scroll, and need a pause between scrolls | | **Enter Text** or a filter action | Changes a page input and may trigger a new result set | Results may appear only after a delay or an AJAX request | The task should continue only after the page reaches the state the next action expects. Use the smallest setting that gives the page enough time to reach that state. ## Choose the right control | If you see | Use | What it changes | | - | - | - | | Content appears a moment after a page opens or a button is clicked | **Wait before or after an action** | Pauses the task for a fixed or configured period before it continues | | The task must not continue until a specific element appears | **Wait for an element or XPath condition** | Holds the next action until the expected page state is present | | A click starts a background request and results appear later | **AJAX or network-request wait** | Gives the background request time to finish before extraction continues | | A page or request needs more time than usual | **Timeout** | Sets how long Octoparse waits before treating the action as timed out; it does not make the page load faster | | Go to Web Page, Click Item, or Click to Paginate fails occasionally | **Retry** | Runs the same supported action again when its retry condition is met | | An infinite-scroll list stops adding items | **Scroll interval and stop condition** | Spaces out scrolling and stops when the page no longer returns new content | ## What a retry does Retries are available for **Go to Web Page**, **Click Item**, and **Click to Paginate**. A retry starts that same action again; it does not continue from the point where the previous action stopped. Set three things for a retry: the condition that triggers it, the maximum number of attempts, and the wait between attempts. Depending on the retry settings, Octoparse can also change the proxy or User-Agent, or clear browser state before trying again. These are optional changes and can end a logged-in session. ## Test a change 1. Find the first action that does not reach the expected page state in the event log or a small Local test. 2. Change one control only: a wait, a timeout, a scroll setting, or a retry. 3. Run the same small sample again. 4. Check the screenshot, final page, and several output rows. An action completing does not by itself mean the data is complete. Longer waits and repeated actions can increase runtime and may produce duplicate records. Stop a run when retries repeatedly return an access page or no new valid data. This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.