For AI agents: visit https://help.zluri.com/llms.txt for an index of all pages formatted in Markdown and endpoints in OpenAPI. Append .md to any documentation page URL to get its markdown version.
Double-check your data against the schema every time.
Use the /entities/schemas API to fetch the latest field definitions.
If validation fails, check the error message and fix the fields accordingly.
Send data in a paginated manner of 1,000 records or fewer — this keeps things fast and reliable.
Use a unique pageNumber for each upload. Reusing a number will overwrite previous data.
Group your data logically — users first, then groups, group-users, and so on.
Retry failed requests with exponential backoff (especially for 5xx errors or timeouts).
Always log the error responses — they help pinpoint what went wrong.
If a sync is off-track, cancel it early instead of waiting till the end.
Python
# Sample retry logic
for attempt in range(max_retries):
try:
return client.upload_snapshot_data(sync_id, entity, data, page_number)
except Exception as e:
if attempt == max_retries - 1:
raise
time.sleep(2 ** attempt) # Exponential backoff
Don’t start a new sync on the same instance until the previous one is complete.
Wait at least 6 hours between full syncs to avoid throttling.
If you hit a 429 Too Many Requests, follow the retry-after header or queue the request.
Add your email to notificationEmails while creating the instance to get sync updates.
Use the Sync Status API to check if a sync is running, failed, or finished.
Keep logs of what you uploaded and which pages succeeded or failed.
Follow the right order: users → groups → group_users → activities.
Always finish your sync. Data won’t show up in Zluri until the sync is completed properly.
