Cloud Lesson｜Want to realize automatic resource backup? After reading this cheat book, no longer banana green~
Reading Fengyun is a master of Huawei cloud information. He is good at diversifying the presentation of complex information. There is always a picture (Yuntushuo), a simple blog post (cloud small lesson) or short video (cloud video hall) that it produces. You quickly get started with HUAWEI CLOUD. For more exciting content, please click here
Abstract: When sets the "alarm" to the backup system, sometimes it may not perform as expected. This article will take you to understand how to implement automatic resource backup in HUAWEI CLOUD backup service, freeing your hands~
This article is shared from Huawei Cloud Community " [Cloud Small Lesson] Basic Services Lesson 87 Want to realize automatic resource backup? , no longer banana green ~ 16130767c80114 ", author: Reading Fengyun.
Huawei Cloud Backup and Recovery (CBR) can perform manual or automatic backup for cloud servers ECS, BMS, cloud hard disk EVS, file system SFS Turbo, and IDC data center file sets, databases, etc. manually or automatically. The function is very powerful.
Some small partners set an "alarm clock" for the backup system when using the cloud backup process, hoping that the system can automatically back up resources at the set point in time, and automatically delete expired backups. However, in actual use, the system sometimes does not seem to perform as we imagined.
This "alarm clock" doesn't listen to orders, it's so green~ what should I do? How to solve it?
Today, the editor will take you into a small class of backup strategy, a lesson package ~ from entry to proficiency~
Set the appropriate storage capacity
Although this lesson introduces how to set up a strategy, if you want the strategy to work normally, then the appropriate storage capacity is also very important!
Sometimes, the policy setting is no problem, but because the storage library capacity is too small, the storage library cannot fit the backup, which will cause the policy to not be executed as expected.
So, how to set our storage capacity according to our strategy?
At this time, I took out the formula for calculating the storage capacity of the small book:
Repository capacity (GB) = (disk capacity (GB) + backup retention time (days) / backup period (days) disk daily change data volume (GB)) 120%
Description of is to keep by the number of copies, it can also be calculated by converting to the retention time. For example, if a company performs backups once a day and the number of reserved copies is set to 7, it can be converted into a retention period of 7 days.
For example: If a financial company A has an 800GB cloud server, it has already used 200GB. When calculating, you need to use the disk capacity instead of the used amount, which is 800GB. The daily data change is about 10GB. The company's planned backup strategy is to perform 2 backups every day at 02:00 am and 20:00 at night, and the backup retention period is 1 month. Then the size of the cloud server backup repository purchased by the company can be estimated by the formula:
Storage capacity = (800+30/(1/2) 10) 120%=1680GB
What should I do if I have bought a storage library and found that the capacity is not enough after reading the formula? It doesn't matter, expand the storage library with one click! However, it does not currently support shrinking the storage library. Don't expand it too much~ If you really want to shrink, you can buy a new small storage library, and it will be a lot to migrate the backup resources in the past~
Appropriate storage capacity can hold all the backups generated by policy execution~ If the storage library size is not appropriate, backup execution may fail, backups are not deleted according to the retention rules, etc.
Setting the appropriate storage capacity is one of the first conditions for the normal operation of the strategy~
Setting a strategy is like setting an alarm clock. You can mainly refer to the strategy parameter table to complete the setting.
Backup cycle: Set the day to back up, you can back up every day of the week or every other day. Here you can see that the editor’s backup rule is set to backup once every Monday, Thursday, and Sunday at 11 pm, then 3 backups will be generated a week.
Backup time: several times on the day of the backup, here I set the backup at 23:00, currently only the whole point can be set~
The more frequent the backup, the more backups are generated, and the larger the storage library capacity required.
Retention type: You can choose to keep by quantity, by time or forever.
Rule details: Because of the backup rules set above, only 3 backups a week, the editor only wants to keep the backups generated in a week, and the retention rules are only set to keep 3 backups.
Advanced options: You can set daily backup rules, etc. The editor here sets up a weekly backup rule to ensure that the latest backup generated every week will be retained on the basis of keeping 3 backups a week. Since the advanced options and reserved types are not conflicting, you can set both at the same time.
If today is the 30th, the implementation of this strategy is as follows:
The date marked with the time is the date when the backup was generated.
The backup in gray time has been deleted.
The green time backup is kept.
If no weekly backup rules are set, only the 25th, 26th, and 29th backups will be kept.
But also need to pay attention to the following points:
- It is recommended to set the time more loosely, otherwise there may be a risk of backup skipping.
16130767c80357 Description: The backup time point set by the backup policy of a disk At 00:00, the disk starts to back up. Due to the large incremental data backed up by the disk this time, or there are many backup tasks performed at the same time during this time period, the backup task takes 90 minutes and the backup is completed at 01:30 . Then the backup time of 01:00 will be skipped, and if the backup is performed at 02:00, only two backups will be generated.
- When performing the backup time and the time of copying the backup, try to ensure that the backup strategy is executed and the backup task is completed before executing the copy strategy, otherwise the copy backup may fail.
- It will not cause any performance impact on backup resources. However, in order to ensure the integrity of the backup data as much as possible, it is best to set a time in the early morning and no data is written to the disk for backup.
After setting the policy, don’t forget to bind it to the repository that you want to automatically back up~ The binding is successful, remember to make sure that the "alarm" is in the "on" state!
Under normal circumstances, after setting the appropriate storage capacity and strategy, the "alarm clock" will automatically operate according to the settings, without manual intervention.
Of course, accidents happen occasionally. for example:
- If the retention rules are modified during the execution of the policy, the backup may not be automatically deleted. For specific reasons, please refer to the common question of cloud backup "Why does the retention policy not take effect after modification?"
- Forgot to bind the policy to the repository, forget to open the policy...
- The repository is bound to new resources, the automatic binding is turned on, and the repository is forgotten to expand...
Don't worry, let's find the root cause, see what happens!
Because the storage library capacity is insufficient and the automatic backup fails, we can expand the storage library, or adjust the frequency of the strategy to reduce the number of resources bound to the storage library...
Before automatic backup, check the opening and binding of the strategy...
If you modify the retention rules, you can manually delete the backups generated by the old policy...
If you encounter any unexpected situation, please feel free to spit out with the editor in the comment area~
Get knowledge about cloud backup, welcome to poke here~
Click to follow to learn about Huawei Cloud's fresh technology for the first time~ ]( https://bbs.huaweicloud.com/blogs?utm_source=segmentfault&utm_medium=bbs-ex&utm_campaign=other&utm_content=content)