- 最新
- 投票最多
- 评论最多
The use of daily, weekly, and monthly backup rules with different retention periods provides several important advantages over using a single daily backup rule with a long retention period:
-
Storage cost optimization: Different retention periods allow you to keep fewer backups as time passes. For example, you might keep daily backups for 5 weeks, but only monthly backups for 5 years. This tiered approach significantly reduces storage costs compared to keeping every daily backup for 5 years.
-
Granular recovery options: This approach provides different levels of recovery granularity based on how recent the data is. You have daily recovery points for recent data when you need fine-grained recovery options, while still maintaining longer-term recovery points at wider intervals.
-
Organizational policy alignment: Multiple backup rules with different frequencies and retention periods make it easier to implement organization-wide backup policies that align with different compliance and business requirements. For example, some data might need daily backups for operational recovery, while other data requires longer retention for compliance.
-
Resource tagging efficiency: You can use tags to assign resources to specific backup plans. For instance, you could tag resources with "backup:daily" or "backup:monthly" to include them in the appropriate backup plan, making it easier to manage which resources get backed up at which frequency.
-
Backup management simplification: Having separate rules makes it easier to manage and audit your backup strategy, as you can clearly see which backups are being retained for which purposes.
AWS Backup handles overlapping rules efficiently. When rules overlap, AWS Backup retains the backup under the rule with the longer retention period, avoiding unnecessary duplication of backup data while ensuring your retention requirements are met.
Sources
Backup plan options and configuration - AWS Backup
Backup and recovery using AWS Backup - AWS Prescriptive Guidance
AWS Backup - S3 Incremental Backup and Potential Duplication | AWS re:Post
How continuity management works - AMS Advanced User Guide
The main reason is cost and practicality.
With daily backups only - you need to pay for 365*5 = ~20k snapshots With daily, weekly, monthly, yearly backup rule - you need to pay for = 100 snapshot.
Yes, backups are incremental but it will still cost something. Also, for practicality, it does not make sense to store daily backups from 4-5 yrs ago. Does it? Many of the times, the backups are kept for compliance reasons, and not necessary to keep daily.
Hi there, here is the break down: Let's keep in mind they all help in term of data lost. Overall it depends on your infrastructure and/or how you would like to design your Backups.
- Daily Backup rule: Runs once a day, leveraging the hourly lost of data of the previous Day( in case that happens). >> 7 snapshots in a week
- Weekly Backup rule: Runs once a week, will help to restore your data as it was at the time of that Day of week. >> ~ 4 snapshots in a month(30 days)
- Monthly Backup rule: Runs once a month, will help you restore your lost data as it was on the last day of the month. >> 12 snapshots in a year.
- Yearly Backup rule: Runs once per year.
When it comes to disaster recovery, the time taken to restore(how long it takes) your data is also influenced by the size of the data to be restored(how big is your backup). Plus other factors such as disk size, network etc. We can easily conclude the fastest restorable snapshots will be the daily ones. Because they store less data than the others. On the other hand, you can target a restore based of a specific month. For example you can restore your s3 buckets(Full or Partial) to be like it was at the month of January etc. Since you're deleting your daily and weekly backup after few days for cost effectiveness for example. You can also only copy the monthly or yearly snapshots in a different region/account to reduce cross-region charges on a daily or weekly copy. Retaining your daily snapshots for 5 years will accumulate in your vault and that might be pain to search it for a restore. By default, a vault can show 100 recovery points per page. Having 1 resource backed up and retained for 5 years ~= 1800 recovery-points in a vault. Multiply this number by the resources you would be backing up.
Hope this clarifies.
Thanks
相关内容
- AWS 官方已更新 2 个月前
