| 开发者 | epicshift |
|---|---|
| 更新时间 | 2026年9月4日 02:16 |
| PHP版本: | 7.2 及以上 |
| WordPress版本: | 7.1 |
| 版权: | GPLv2 or later |
| 版权网址: | 版权信息 |
No. Every check runs against your own site. That is the point: you can find out what is wrong before involving anyone.
By checking, rather than by trusting us. It asks for no keys, no account and no credentials of any kind. It contacts no external service: the only request it makes is to your own domain. It writes nothing to your products, your orders or your settings. The source is plain, commented PHP with no build step and no minified files, so you or your developer can read every query it runs before you run it. And where a check cannot prove something, it says so on the page instead of guessing.
No. It tells you what is wrong and what the fix is. The fixes are settings changes, and most of them take a minute; a few need your host.
It writes nothing to your content. It reads settings, PHP configuration, product metadata and order line items, and it requests your own REST API root. It keeps two temporary things and three tiny permanent ones, and deactivating the plugin deletes all of them. The temporary ones are WordPress transients: a cached count of stranded orders, so the dashboard warning does not run a database query on every admin page (expires within the hour), and a record that you clicked "Hide for a day" on that warning (expires in a day). The permanent ones exist so the plugin can ask you for a review politely instead of immediately: the date you first ran it, how many times it has run, and your answer if you dismissed the ask. No content, no orders and no settings of yours are ever written to.
Two possibilities. Your integration may collect orders in a way this check cannot see, in which case the finding is wrong and worth telling us about. Or the order was collected in the last day, which is inside the grace period the check allows before it reports anything, and something else changed its date. The check errs towards silence: an order is only reported when it is still waiting to be fulfilled, was placed after the last recorded visit, and has been waiting more than 24 hours. Fresh orders are never reported, because suppliers poll on their own schedule. One more thing it cannot separate: an order containing a supplier product that you decided to make or ship yourself looks identical from here.
No, and it says so on the page. It examines the 200 oldest orders in the window and names the 20 oldest of those, because two hundred bullet points is a wall rather than a finding. When it stops at either limit it prints a line saying the total is a floor and not a count. A diagnostic that quietly stops counting reads exactly like one that found everything, which is worse than useless.
It means the integration looked at the order and found no product in it that it recognises as one of its own. There are four causes, and they are worth checking in this order: the product was never linked to the supplier's copy of it, the link was lost when a sync failed part way, the product was created directly in the shop and the supplier has never heard of it, or the integration cannot reach your shop at all. The REST API, permalink and leftover checks in this plugin cover the last two. The full walkthrough, with how to confirm each cause, is at orderguard.dev/fix/order-was-not-imported-printful.
A 404 during sync means the supplier's servers called into your shop and your shop answered "no such page". The usual causes are permalinks set to Plain, a security or firewall plugin standing in front of the REST API, or hosting that cannot answer a second request while it is building a page. Checks 2, 3 and 4 measure exactly these, and the guide at orderguard.dev/fix/printful-sync-not-working-rest-api-404 explains how to tell them apart.
That error appears at checkout when no shipping method answers for the customer's address, and with print-on-demand it usually means the integration failed to return live rates at that moment. It is the most expensive error on this list, because the customer is standing at the till and leaves. The two-minute protection is a flat-rate fallback in the shipping zone, and the guide at orderguard.dev/fix/no-shipping-options-available-print-on-demand walks through it.
Then the cause is not in your hosting or your catalogue — it is at the supplier or inside the integration plugin. That is worth knowing: it means the time you would have spent changing settings is time you can spend on a support thread instead. Copy the report from the bottom of the page and paste it there.
No, and it is worth reading carefully. It means your server did not answer its own request within eight seconds, and there are two possible causes: something is blocking the request, or your hosting has too few PHP processes to answer a second request while it is still working on the first. Both of them break sync, so neither is a false alarm. The check tells you how to tell them apart.
Partly. The REST API, permalink, blocking-plugin and PHP-limit checks all run. The order and catalogue checks need WooCommerce and will say so.
Yes. Topics on this plugin are read. No promise is made about how fast, because a promise about response time is one we would eventually break, and that is worse than not making it.