Troubleshooting

When something looks wrong, start with the Logs page — most issues surface there with a precise error message you can act on. Below are fixes for the rest, grouped by where you noticed the problem.

Tips

Almost every issue below has a fix you can apply from the WordPress admin without touching code. Where a host-level change is the only real fix, it's called out in a callout.

Connection errors

These show up on the Connection tab when you click Test & Save.

Symptom	What it means	What to do
HTTP 401	Wrong credentials	Open UnoPim's API panel, double-check the client ID, secret, username and password, then re-paste all five fields and click Test & Save again. Generate a fresh API client in UnoPim if needed.
HTTP 0	The site can't reach UnoPim at all	Confirm the UnoPim URL works in a browser. If your site has external requests blocked at the host level, ask your host or developer to allow your UnoPim domain.
HTTP 404	Wrong base URL	Use only the bare admin URL, e.g. `https://pim.example.com`. Don't add `/api/v1/rest` or any trailing path — the connector adds those itself.
Saves OK, but Channels & Stores is empty	The UnoPim user can sign in but isn't allowed to read channels	In UnoPim, give that user a role that includes channel read permission, then come back and click Refresh channels.
Connection drops mid-sync	Session expired and couldn't refresh	The connector retries automatically. If errors continue, open the Connection tab and click Test & Save again to force a fresh sign-in.
Warning about missing security keys	Your site doesn't have its standard WordPress security salts set	Ask your developer or host to add WordPress secret keys to your site config. Until then, credentials are stored in a less-secure form.

Sync failures

A "Pull Full" never finishes

Something in the queue has stalled.

Open the Logs page and look for recent rows with error status — they usually point at a specific entity (a category, a product, an attribute) that's blocking the queue.
Fix the underlying issue (for example, a missing family in UnoPim) and click Pull Full again.
If nothing has run for more than 10 minutes and Logs is quiet, the background queue may not be firing — see Cron not firing below.

Some entities pull but others don't

Each entity type (categories, attributes, products, variants…) is pulled independently, so one type can fail while others succeed. Open Logs, filter by status = error, and read the message.

What you see in Logs	Likely cause	Fix
Categories pull, products don't	Products are being skipped, or no default family is set	Check the Core Fields tab and make sure a Default family is selected.
Products pull, variants don't	The product's family in UnoPim doesn't define the variant axes correctly	Edit the family in UnoPim, then click Pull Full.
Attributes pull, options don't	The UnoPim user can't read attribute options	Update the user's role in UnoPim to include attribute-option read permission.

Push silently does nothing

When you save a product in WooCommerce and nothing reaches UnoPim, check these in order:

Open Channels & Stores and make sure Push is toggled on for the channel you expect.
If you just pulled the same product, the connector waits a short cooldown to avoid ping-ponging the same change back. Wait a minute and try again, or lower the Debounce seconds value.
Open Logs and look for a row about that product. The message will tell you exactly which gate stopped the push.

Push returns HTTP 422

UnoPim received the data but rejected it.

Open the Logs row.
Read the message — it usually names the offending attribute.

Message hint	What it means	Fix
"attribute X not in family schema"	Your mapping points at an attribute that isn't part of this product's family in UnoPim	Either edit the family in UnoPim to include the attribute, or remove that attribute from your mapping.
"value type mismatch"	You mapped a text field to a numeric attribute (or similar)	Change the WooCommerce field in the mapping to one of the same type.
"missing required field"	The product is missing a value the family marks as required	Either fill the value in WooCommerce, or remove the "required" mark on that attribute in UnoPim.

Push returns HTTP 401 mid-batch

The session expired during a long batch. The next push retries automatically with a refreshed sign-in. If it keeps happening, ask UnoPim's administrator to extend the API session lifetime.

Push returns HTTP 500

Something went wrong on the UnoPim side — not the connector. Check UnoPim's own logs. The connector will retry automatically with a back-off delay.

Cron not firing

WordPress runs scheduled tasks (like delta pulls every 5 or 15 minutes) when visitors hit your site. On low-traffic stores, those tasks queue up and fire late or not at all.

Warning

If your scheduled pulls always run hours late — or only when you click around the admin — your site needs a real "system cron" set up by your host. This is the only fix that requires a change outside the WordPress admin.

Ask your host or developer to set up a system cron that triggers your site every minute. Most managed WooCommerce hosts have a one-click option for this.

After system cron is set up, scheduled pulls fire on time regardless of visitor traffic.

Mapping issues

Auto-map matched the wrong attributes

Auto-map guesses based on similar names and codes. If two UnoPim attributes look alike, it may pick the wrong one.

On the affected scope (channel + locale), click Clear mapping.
Use + Add WooCommerce field to set each row by hand.
If you have many channels with the same correct setup, save it as a preset and apply it elsewhere.

Mapping wizard is slow

Large UnoPim catalogues with thousands of attributes can make the mapping screen feel sluggish.

Use the search box to narrow down attributes before opening dropdowns.
Reduce the per-page count on the Attributes tab so fewer rows render at once.
If it's still slow on huge catalogues, your developer can extend the search and rendering limits.

The connector only lists fields that already have at least one stored value. If your field is brand new and empty:

On the Attributes tab, click + Add custom column.
Type the exact field key.
It now appears in the dropdown like any built-in field.

"Refresh targets" does nothing

Your sign-in to UnoPim probably expired between clicks.

Open the Connection tab and click Test & Save.
Try Refresh targets again.

If targets are still empty, the UnoPim user's permissions may have been reduced — verify in UnoPim.

Variant issues

Symptom	Cause	Fix
Variations land without variant attributes	A variant axis attribute is configured as plain metadata instead of a taxonomy	Open the Attribute Mapping tab and click Refresh targets, then Pull Full.
Variations push without their parent	Variants got pushed individually before the parent was ready	Use Push All Products instead of pushing variants one by one. The connector handles parent-then-variant order for you.
"Axis outside family schema" error	The product's family in UnoPim doesn't allow that attribute as a variant axis	Edit the family in UnoPim and add the attribute as a variant axis, then push again.

Performance issues

The Logs page is slow

The log table has grown large.

Click Flush logs → older than 7 days at the top of the Logs page.
Lower the per-page count to 25 while you're investigating.

Symptom	Cause	Fix
Translations created but empty	Source language pulled before the target language attributes were ready	Run Pull Full once more after the attribute schema is stable.
WPML language column shows the wrong language	Locale-to-language mapping mismatch	Standard locales like `en_US` → `en` and `fr_FR` → `fr` work out of the box. For unusual cases (regional variants), the connector also supports a custom locale-to-language map if your developer wants to extend it.
All locales merge into one product	WPML isn't active or detected	Make sure WPML is installed and active. If you use a different translation plugin, your developer can build a bridge to it.

See Multilingual for the full picture.

Data setup issues

Symptom	What to do
"Table doesn't exist" errors after upgrade	Deactivate and reactivate the connector from Plugins → Installed Plugins. This re-runs the setup. If the error persists, ask your host to ensure the WordPress database user has create/alter table permissions.
Schema didn't migrate after a plugin update	Deactivate then reactivate the connector once. Migrations re-run on activation.
The log table is enormous	Set Log retention to 7 or 14 days in General Settings, then click Flush logs → older than 7 days once. Going forward the connector trims logs automatically.

Multisite issues

Symptom	Cause	Fix
Pull writes to the wrong site	The channel-to-site mapping is wrong	Open Channels & Stores on the relevant site and update the Site dropdown for that channel.
The connector isn't activated on some sites	Network activate skipped a site	Visit each site in My Sites, open Plugins, and activate the connector individually.
Cron only fires on the main site	Each site in a network has its own schedule	Ask your host to set up a system cron that ticks every site, not just the network's main one.

See Multisite for the full picture.

How to file a useful bug report

If none of the above resolves your issue, open a ticket with:

Plugin version — visible on the Plugins page.
WordPress, WooCommerce, and PHP versions — from Tools → Site Health → Info.
What you did — the exact button/click sequence.
What you expected vs. what happened.
5–10 rows from the Logs page around the time of the issue (use the time filter).
Screenshots of the failing screen, if any.

The faster the support team can reproduce, the faster they can fix it.