架构师实录：系统稳定性的保障与思量

后端

2024-01-30 05:05:49

打造稳定可靠的系统：从架构、技术到运维的全面指南

随着业务发展，企业对系统的要求也在不断变化。从高可用性到可扩展性，再到可靠性和安全性，每个阶段都需要我们针对性地优化系统架构、选择技术和实施运维措施。

一、架构层面

在系统架构设计中，我们要权衡不同阶段的考察指标，作出合理的取舍。

初期阶段： 优先考虑高可用性。采用分布式架构、增加备用实例和数据冗余，保证系统故障时的快速恢复。

成长阶段： 注重系统性能和可扩展性。采用负载均衡、分布式缓存技术提升性能；微服务和云原生架构增强可扩展性。

成熟阶段： 强调系统可靠性和安全性。通过故障恢复、数据备份手段提升可靠性；采用身份验证、授权、加密措施保障安全性。

二、技术层面

技术选型要综合考虑以下因素：

可靠性： 采用负载均衡、分布式缓存、热备等技术提升系统可靠性。

import com.google.cloud.storage.*;
BlobId blobId = BlobId.of("my-bucket", "my-object");
BlobInfo.Metadata metadata =
    BlobInfo.Metadata.newBuilder()
        .setContentType("text/plain")
        .setContentEncoding("utf-8")
        .build();
BlobInfo blobInfo = BlobInfo.newBuilder(blobId).setMetadata(metadata).build();
Storage storage = StorageOptions.getDefaultInstance().getService();
Blob blob = storage.create(blobInfo, "Hello, world!".getBytes(StandardCharsets.UTF_8));
System.out.println(
    "Blob " + blob.getBlobId() + " uploaded to bucket named " + blob.getBucket());

性能： 利用分布式数据库、内存缓存、异步处理技术提升系统性能。

from google.cloud import bigquery

# Construct a BigQuery client object.
client = bigquery.Client()

# TODO(developer): Set table_id to the ID of the table to create.
# table_id = "your-project.your_dataset.your_table_name"

schema = [
    bigquery.SchemaField("name", "STRING"),
    bigquery.SchemaField("post_abbr", "STRING"),
]

table = bigquery.Table(table_id, schema=schema)
table = client.create_table(table)  # API request

print(
    "Created table {}.{}.{}".format(table.project, table.dataset_id, table.table_id)
)

可扩展性： 采用微服务、云原生技术实现系统弹性伸缩。

import (
	"context"
	"fmt"
	"io"

	"github.com/GoogleCloudPlatform/golang-samples/run/helloworld/pkg/run"
)

// HelloPubSub is an example of handling a Pub/Sub message.
func HelloPubSub(ctx context.Context, w io.Writer, e run.Event) error {
	var pubSubMessage struct {
		Data []byte `json:"data"`
	}
	if err := e.Data(&pubSubMessage); err != nil {
		return fmt.Errorf("e.Data: %v", err)
	}

	fmt.Fprintf(w, "Message: %s", string(pubSubMessage.Data))
	return nil
}

三、运维层面

系统上线后，持续的运维工作至关重要。

监控： 收集和分析日志、指标数据，及时发现系统问题。

# A simple Google Cloud function to check if there's a custom header in the
# current request.
runtime: go113
entrypoint: check
functions:
  check:
    # Set a custom header when you make your request
    # e.g. using curl with -H flag: curl -H "x-goog-meta-test:my-header"
    # https://{your_cloud_function_url}
    https:
      allow_http: true
      body: "*"
      logging:
        level: debug
    env_variables:
      HEADER_TO_CHECK: x-goog-meta-test
      HEADER_VALUE_TO_CHECK: my-header

备份： 定期备份数据，防止数据丢失。

        // Make an explicit request for the IAMPolicy object.
        var iamPolicy = await folder.GetIamPolicyAsync();
        // Retrieve the existing bindings in the IAM policy.
        var policyBindings = iamPolicy.Bindings;
        // Create a new IAMPolicy object with the desired bindings.
        var newBindings = new List<Policy.Types.Binding>();
        foreach (var policyBinding in policyBindings)
        {
            newBindings.Add(
                new Policy.Types.Binding
                {
                    Members = policyBinding.Members,
                    Role = policyBinding.Role,
                    Condition = policyBinding.Condition
                });
        }
        newBindings.Add(
            new Policy.Types.Binding
            {
                Members = new List<string> { "domain:google.com" },
                Role = "roles/resourcemanager.projectViewer"
            });
        var newIamPolicy = new Policy
        {
            Bindings = { newBindings }
        };
        // Set the IAMPolicy with the desired bindings.
        await folder.SetIamPolicyAsync(newIamPolicy);
        Console.WriteLine($"New IAMPolicy was set for {folder.FullName}.");