[DOCS-12058] Atlas docs should warn that a restore will require an application restart Created: 14/Sep/18  Updated: 29/Oct/23  Resolved: 05/Dec/18

Status: Closed
Project: Documentation
Component/s: Atlas
Affects Version/s: None
Fix Version/s: None

Type: Improvement Priority: Major - P3
Reporter: Amit Jain Assignee: Anthony Sansone (Inactive)
Resolution: Fixed Votes: 3
Labels: atlas
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Issue Links:
Related
Participants:
Days since reply: 5 years, 21 weeks, 5 days ago
Epic Link: DOCSP-3127
Story Points: 0.3

 Description   

Description

If we look at the docs for restoring a replica set in Ops Manager, it has a note saying:

You must ensure that the MongoDB deployment does not receive client requests during restoration.

Because similar notes are missing from the Atlas docs, users may mistakenly believe they can keep their application seamlessly online while doing a restore.

 

On the Atlas pages for restoring from backup, please add 2 notes:

  • A pre-condition that requests (reads AND writes) to the destination cluster have stopped.
  • A post-condition that the application using the destination cluster should be restarted.

Scope of changes

Modify the existing admonition to include explicit callouts for the consequences of downtime.

For restoration via HTTPS snapshot, we might note the same requirements while the user restores using 

Impact to Other Docs

  • Cloud Manager Restore to Atlas Cluster needs an admonition as well. Consider creating (or cloning) this ticket for the CM portion.

    MVP (Work and Date)

    Resources (Scope or Design Docs, Invision, etc.)



 Comments   
Comment by Ravind Kumar (Inactive) [ 14/Sep/18 ]

The Continuous Backup Restoration and the Cloud Provider Snapshot Restoration sections do mention that the target cluster requires downtime. However, there's no harm in being explicit the consequences of downtime - mainly that any reads/writes to the target cluster must stop for the duration of the restoration procedure, and the application may need to restart and/or update its connection string (e.g. if restoring to a different Atlas cluster). We should confirm w/ engineering as to whether restoration can result in connection string changes if it's the same cluster, or if the restart just allows for refreshing the IP addresses associated to the cluster's DNS hostnames.

Generated at Thu Feb 08 08:04:18 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.