How to tell if you encountered a data related problem
This section explains how to tell if you've encountered a data issue and how to identify it source.
Where the issue comes from?
How to tell if you've encountered a data related issue?
In order to facilitate the comprehension and indentification of your problem, here is the Data Flow Model and five areas of where your problem can come from:
Apostrophe Project (bug / regression)
Datastrem - Apostrophe connection
Data provided by the business
Import into Global DB (import rules)
Export to Datastream (export rules)
1. Apostrophe Project
The most frequent data related issue is missing information: not all CAIs are available, products miss their labels ("Coming soon" or "New"), some information is missing from the SKU table or there is no SKU table at all, a dealer do not have opening hours. All of these seem to be a data issue, but in the fact 90% of those are not.
First thing to check, when we have a missing detail are Apostrophe Pieces:
If the information is present in the Apostrophe pieces, it means that the problem is platform related.
For the problems listed above:
not all CAIs are available - you need to check if the CAI in question is present in Tyres / Articles menu. If no, go to point 3 - Business provides incorrect Data. If yes, check its start/end date are correct, if yes it means it's an Apostrophe problem.
products miss their labels ("Coming soon" or "New") - in the Tyres / Products menu check if the chosen Product has Soft Launch / Start Date and if they are correct. If yes it is an Apostrophe issue. Here is an example of this type of issue.
some information is missing from the SKU table - check if the information is present in the Tyres / Articles menu. If yes, it's an Apostrophe issue.
there is no SKU - in the Tyres / Products menu, in Technical tab check if the chosen Product has at least one article listed. If yes, it's an Apostrophe issue. Here is an example of this type of issue.
dealer do not have opening hours - in Dealers / Dealer menu look for a dealer in question, check if it has a Google Place ID. If yes, it's an Apostrophe issue.
2. Datastream - Apostrophe connection
If you cannot find the latest data in the pre-production and you are sure Lizeo has already pushed the update, it is possible that you're facing a connection issue and feed is simply not going through.
Datastream module is working using queue management - Datastream informs all the systems that new data is available and each system fetch the data when ready.
What's more, in queue management feeds are being processed one by one in the order they arrive. It means that if a lot of feeds was pushed in the same time, the processing time can be quite long.
Two checks can be made on the connection level:
In the pre-production in Configurations / Subscriptions menu, you can check the latest feed information on your own. If you have the feed ID provided by Lizeo, you can compare if is a match. If not, you can refresh the subscription. The detailed information on how to do so can be found in the Apostrophe - data information chapter.
Ask Kamila Jamroz from Data Team to check if she can see the latest feed/data in the Datastream. If yes, there are two possible explanations:
Your feed is stuck in the queue. The solution is to ask a back-end developer (Harouna Traore or Quentin Mouraret) to verify where your feed stands in the queue.
Connection between the project/Datastream is not working correctly (more unlikely). This problem requires a Git ticket and back-end checks.
3. Business provides incorrect Data
The quality of the data we receive from Business is a separate issue and it varies from one Business Line to another or even between the different imports of the same file (for example, Google Place ID deletion, when country forgets to download the latest file version from Lizeo's Back Office).
Let's go back to the first problem from Apostrophe Project step. You are unable to find a CAI/a Product in Apostrophe. The next step is to check the Business entry file (Product Catalog or Product Template). If the element you're looking for is not available in the entry file, you need to contact the person who provided the file and ask to correct their data.
Where to find latest information?
Latest dealer information can be checked or the complete file can be downloaded from the Lizeo BO.
Product Template can be downloaded from the Lizeo BO.
Vehicle information can be checked in the Lizeo BO (for TSOv3 users) or in TSU.
Latest catalogues can be found in Jira tickets or in Lizeo BO (does not apply to AMN - for more information please check in dedicated project FAQ).
4. Global DB import (import rules)
This check is strictly related to the previous one, as it requires the same verifications in the Lizeo Back Office.
What to does it imply if you cannot find the information you are looking for?
Files have not been treated yet - Lizeo hasn't finish the import. The best thing to do is to monitor Jira ticket containing the file you're interested in. Once the import done, the ticket will be updated and you'll be able to make your checks. In case of any problems, you can contact Séverine Ramadour from Lizeo.
Some data feeds has not only import rules, but also export rules. It means that certain informations are being added on the Global DB export, when pushed to the Datastream. In some cases, Lizeo's Back Office is not reflecting the reality and the full data can only be seen in the Datastream (or on the website).
5. Datastream export (export rules)
As mentioned, date is not being simply pushed to the Datastream. There are export rules attached to the Global DB / Salsify outbound to make data compliant with Business requests. This is why the Lizeo's Back Office does not always reflect the data within Business databases and websites.
Some of this rules has been already documented within Data Gitbook chanel (you can find more information in the DATA MODELS and in each project respectively), this part is still in progress. Some rules are impossible to retrieve as they quite old and not documented at the time.
If you checked all the other steps and you are sure that the error is caused by one of the rules a Jira/Git ticket needs to be created in order to alert the Data Team and Séverine Ramadour.
Last updated
Was this helpful?