Technical Teams, Customer Success Specialists, Brand's POS programmer
Description of issue:
The failed order report is showing that orders failed with the reason POSisoffline but the POS was online with Brink the entire time.
When we communicate with Brink we send the orders to the Brink AWS server. This server then sends the order to the store and when the store responds back to the Brink AWS server Brink returns us a success or failure for the order. POSisoffline is the response that Brink sends to us anytime that the Brink AWS server sends an order to the store and does not get a response within their timeout window. This can make the issue hard to troubleshoot because we do not have visibility in to the details of what happened at the store but in most cases one of the issues below is what causes this issue to occur.
Solutions to review:
Internet connectivity and white listing:
The most common cause for this error is that the network blocked the message back to the Brink AWS servers. The AWS server is not the same server used for in store orders so even if your white listing rules work fine to allow for in store orders it can still be the culprit for failed Olo orders. The Brink white listing guide can be found here. As a first step be sure to review these rules. The AmazonAWS servers are the most common overlooked URL but we have seen many of these URLs cause the POSifOffline error so please review the entire doc and be sure that you are able to communicate to all of the required URLS from the store.
Time out of sync:
A less common cause for these errors can happen if the time on the master terminal in the store gets out of sync with the Brink AWS server. When the workstation receives the order it will not respond if the order is considered stale which happens after 10 seconds. Therefore if the workstation is a few seconds off of the time from the AWS server this can cause some orders to fail. This is often the issue when we see this happen mid day and continue through the end of the day or if you see this error on particularly large orders because they take longer to process. We recommend that you confirm that your system is syncing to internet time so that the time stays accurate. If you would like more details about recommended sync settings please reach out to Brink.
The final cause to check is a network timeout setting. In some systems we have found that there is a time limit to a connection so particularly large orders that take longer to process will get cut off while being sent to the POS. This is pretty rare and will likely require the assistance of your ISP and network administrators to identify
Conclusion and Next Steps:
In the vast majority of cases one of the three issues above have been found as the root cause for chronic issues with the POSisOffline error. There are times when a temporary dip in connectivity or something happening at the store can cause this to occur for 1 or 2 orders. Generally speaking one time order failures will happen occasionally however if this issue is happening regularly or happening for every order over a period of time then it is best to review all of the solutions listed above. If all of these settings are reviewed and you are not able to resolve the issue then reach out to Olo Support at firstname.lastname@example.org and we can contact the Brink API support team to investigate your specific system.