We should provide a way to restart sample bucket load task if it fails for some reason.

1. Create a 5.5 cluster with index nodes in the cluster
2. Start loading the sample bucket into the cluster. Load the travel-sample app since it takes longer for the load to complete.
3. While the task is in progress, start a rebalance of the cluster by adding a node or removing a node. (You will need to use the CLI to do this, since on the UI the rebalance is disabled while the task is in progress. The purpose of rebalance is to cause a failure in sample bucket load task)
4. The rebalance task will cause the bucket load task to fail.

Once the rebalance completes, we see that the sample bucket was created, but since the load task failed, no data would have been created or incomplete number of docs would be created. We should either retry to load the document after failure or let the user retry from UI. Currently, we see that we show the bucket that failed to load on the UI settings, but the user cannot retry the load. Attaching screenshots of the above scenario.