Maintenance FAQ

Between 10am and 11am (BST) on Saturday 19th October Kleene will conduct scheduled maintenance, during which time the Kleene platform will be unavailable.

What are we doing?

We are optimising our database server which would requires us to restart our system. This will impact the availability of the Kleene app and services for 1 hour on Saturday 19th October between 10am and 11am BST).

During the database restart process, it will be necessary for us to temporarily stop our current systems in order to take backups before restarting the database server. We understand the importance of maintaining the highest standards of reliability, and we appreciate your understanding and cooperation as we work to implement these enhancements.

When will this happen?

Between 10am and 11am on Saturday 19th October 2024 (BST)

What do I need to do?

There's no action required from Kleene customers at this moment.

Will there be any downtime?

The maintenance window is between 10am and 11am (BST) on Saturday 19th October.
This means that Kleene will be inactive during this time.

You will not be able to log into your Kleene account during the maintenance window.

What to expect during the database restart?

It is expected during the restart that any scheduled extracts and transforms will not execute. Any extracts or transform that are scheduled to start before 10am, but continue running into the maintenance window will be affected. However all extracts and transforms should run on their next existing schedule after the migration maintenance window finishes.

Note - If it can be avoided, please do not update your existing schedule so that your full pipeline runs before the maintenance window. Doing so may impact the performance of your Kleene system, therefore impacting already scheduled runs.

This database restart does not include any changes to Kleene’s existing features or functionality.

What to expect after restart?

All functionality should work as expected. All scheduled extracts and transforms will run as per their schedule following this migration.

What will happen to extracts that bring in data for one day only?

(This advice is for extracts that load 1 day’s worth of data each day)
Please note that the below advice is based on standard configuration and the outcomes may vary depending on your extract configuration.

• Append
  After missing 1 day of data - 2nd day will extract that day’s data and append to destination table
  1 day will be missing
• Append Incremental
  After missing 1 day of data - 2nd day will extract all data between the last date in destination table to that day’s data and append
  0 days of data missing
• Drop Create Load
  After missing 1 day of data - 2nd day will extract that day’s data and load to fresh destination table
  1 days will be missing
• Truncate Load
  After missing 1 day of data - 2nd day will extract that day’s data, clear the destination table and load into it
  1 day will be missing

What can I do to backfill any missing data?

Backfill tasks will be needed for the missing days

• Check the connector docs to see if backfill window is possible, this should give you specific details on each of your connectors.
• Extract the missing days and insert into the destination table
• Once data is restored, transform runs will generally pick up the missing data and update the pipelines

However there are some Connectors that do not facilitate backfill windows e.g. Google Ads , True-fit-360, TikTok (display)

Transforms that are configured to work only on Saturday 19th Oct 2024 between 10am and 11am will be affected and might need to be run manually to cover the data from Saturday 19th October.

Will my data be safe during the restart?

Yes, data security is our utmost priority. The database stays in the same eu-west-1 region of AWS and there are no changes scheduled to data sources or warehouse.

What happens in case something doesn’t work?

In case of any unforeseen issues during the migration, a roll back plan is in place.

Roll Back Plan for Migration

If there are problems up to 45 minutes into the restart, we will execute a roll back to the previous state.

Roll back plan on the day of migration
Restore the database from backup taken before restart. However any scheduled run missed during the 45mins will still be affected. In the event of roll back , we will capture the results for analysis and work towards a maintenance date in the future.

It should be noted that once we have restarted the database successfully, a full roll-back will not be possible. Any issues that are found as a result of the restart will be resolved on a case-by-case basis. There is a chance that customers may experience data loss as part of missed schedule runs (where backfill is not supported) or system failure that occurs after the migration. The best course of action in this scenario would be to identify the issue and report it to the Kleene support team for resolution.

Who can I contact if I have questions?

Please contact our support team at support.kleene.ai or a member of the Customer Success team if you have any questions or concerns about the migration.