How Do I Set the Number of Concurrent Processes for Linux File-Level Migrations?
Overview
You can set the maximum number of processes that the SMS-Agent can start concurrently (1 to 4) when migrating and synchronizing files of a Linux source server. However, the number of processes that can actually be started depends on how well the source server performs and how many disks and partitions the source server has.
![](https://support.huaweicloud.com/eu/sms_faq/public_sys-resources/caution_3.0-en-us.png)
- Enabling concurrency consumes source server resources, especially disk I/O, bandwidth, and CPU resources. To prevent the services on the source server from being affected, before enabling this function, you can evaluate the impacts on source services or set resource limits for the SMS-Agent.
- If the migration bandwidth from the source server to the target server is less than 300 Mbit/s, you are not advised to enable this function.
Calculating the Maximum Number of Concurrent Processes
The formula is:
Max. concurrent processes = Min. (Disks, Network bandwidth/Disk read or write performance, 4)
where
Disks is the number of disks to be migrated on the source server.
Network bandwidth is the network bandwidth between the source server and the target server. You can calculate the bandwidth by referring to How Do I Test the Network Between the Source and Target Servers Using iPerf?
Disk read or write performance is either the average disk read performance of the source server or the average disk write performance of the target server, whichever is smaller.
Assume that the migration involves four disks, the network bandwidth is 1,000 Mbit/s, the average read performance of the source server is 500 Mbit/s, the average write performance of the target server is 300 Mbit/s. In this example, we need to divide the network bandwidth by the average write performance of the target server (1,000 Mbit/s / 300 Mbit/s ≈ 3). That makes the three values for the formula, respectively, 4, 3 and 4, which means there can be 3 concurrent processes for this migration.
Setting the Maximum Number of Concurrent Processes
On the Configure Basic Settings page, set Migration Method to Linux file-level and Enable Concurrency to Yes, and specify Max. Concurrent Migrations and Max. Concurrent Syncs.
Product Consultation FAQs
- Can I Migrate Physical Servers and VMs from Other Clouds to Huawei Cloud Using SMS?
- Can I Migrate ECSs from Huawei Cloud to On-Premises Environments or Other Clouds Using SMS?
- Can I Ask Huawei Cloud Technical Support to Help Me Migrate Services?
- Does SMS Support Resumable Data Transfer?
- How Do I Migrate an ECS from One Account to Another?
- How Do I Migrate an ECS from One Region to Another Using SMS?
- Will My Services on the Source Server Be Interrupted During Migration?
- How Do I Create an AK/SK Pair for a Huawei Cloud Account?
- How Do I Create an AK/SK Pair for an IAM User?
- How Do I Migrate a Linux Source Server as a Non-root User?
- What Information Does SMS Collect About Source Servers?
- How Do I Select a Target Server?
- How Do I Create an ECS?
- What Is the Block-Level Migration?
- What are Valid Data Blocks?
- How Does Migration Using SMS Differ from Migration Using IMS?
- Why Is the OS Name of the Target ECS Displayed on the ECS Console Different from That of the Source Server?
- How Do I Determine Whether a Migration Is Complete?
- Can I Migrate Only Some Services on the Source Server?
- Can the Target Server Run a Different Type of OS from the Source Server?
- What Determines the Specifications of a Target Server?
- Which Download Option Should I Choose When I Download the Windows Agent?
- Can I Migrate a Source Server If There Is No Corresponding Image Available on Huawei Cloud?
- How Does SMS Differ from IMS in the Migration Scenario?
- How Do I Display the OS Name of a Target Server on the ECS Console?
- What Are the Snapshots of a Target Server Used for?
- In a Linux File-Level Migration, How Do I Exclude Directories from Migration and Synchronization?
- Will an Incremental Synchronization Overwrite Existing Data on a Launched Target Server?
- How Do I Create a Target Server that Meets the SMS Requirements?
- Can SMS Migrate Self-built Databases, Big Data Services, or Websites Deployed on a Source Server?
- Can the AK/SK of a Federated User (Virtual IAM User) Be Used for Authentication During the SMS-Agent Startup?
- How Do I Synchronize Files in Specific Directories After a Linux File-Level Migration Is Complete?
- Will the Migration Be Affected If the SSH Service Does Not Use Default Port 22 on the Source Server?
- How Do I Obtain the SMS Domain Name?
- How Many Resources Will Be Used for a Linux Block-Level Migration?
- How Do I Configure Data Compression for a Linux Block-Level Migration?
- Do I Need to Activate My Windows OS and Paid Software After the Migration?
- Why Does the Target Server Not Synchronize the Clocks on the Source Server?
- How Do I Set the Number of Concurrent Processes for Linux File-Level Migrations?
- Which Directories Are Not Synchronized by Default During Incremental Synchronization on a Linux Server?
- How Do I Verify Data Consistency Between the Source and Target Servers?
- Can I Use a Temporary AK/SK Pair for Migration?
Feedback
Was this page helpful?
Provide feedbackThank you very much for your feedback. We will continue working to improve the documentation.
more