https://www.chenshaowen.com/en/tags/middleware/Recent content in Middleware on Shaowen Chen's WebsiteHugo -- gohugo.ioen©2016 - {year}, All Rights Reserved.Sun, 12 Jul 2026 00:00:00 +0000weeklyhttps://www.chenshaowen.com/en/blog/container-deploy-ceph.htmlSun, 12 Jul 2026 00:00:00 +0000Sun, 12 Jul 2026 00:00:00 +0000https://www.chenshaowen.com/en/blog/container-deploy-ceph.htmlThis article covers single-node validation and three-node production deployment with cephadm. For architecture, operations, and functional testing, see Ceph Architecture and Operations. 1. Disk Preparation An OSD can only use a block device (/dev/sdX, a partition, /dev/loopN). 1.1 Inspecting Available Devices 1 2 3 4 5 6 7 lsblk -o微信公众号MiddlewareContainersCephStorageOperationshttps://www.chenshaowen.com/en/blog/container-deploy-ceph-architecture-operations-and-testing.htmlSat, 11 Jul 2026 00:00:00 +0000Sat, 11 Jul 2026 00:00:00 +0000https://www.chenshaowen.com/en/blog/container-deploy-ceph-architecture-operations-and-testing.htmlCeph is a unified distributed storage system that supports block storage (RBD), object storage (RGW), and file storage (CephFS). The official cephadm can orchestrate each daemon as a container (Docker, or Podman + Containerd); see Deploying Ceph in Containers for details. 1. Architecture 1.1 Topology and Data Paths The diagram below shows a common five-node topology (3 MON + 2 MGR + 5 OSD): MONs only need to come in odd numbers, and in a five-node deployment 3 is enough (tolerating one MON failure); 5 MONs tolerate 2 failures, but the metadata overhead is larger, so there is generally no need to run a MON on every machine.微信公众号MiddlewareContainersCephStorageOperationsArchitecturehttps://www.chenshaowen.com/en/blog/container-deploy-victoriametrics.htmlSat, 11 Jul 2026 00:00:00 +0000Sat, 11 Jul 2026 00:00:00 +0000https://www.chenshaowen.com/en/blog/container-deploy-victoriametrics.htmlVictoriaMetrics (VM for short) is a time-series database compatible with the Prometheus ecosystem. It uses fewer resources than Prometheus and comes in both single-node and cluster editions. The single-node edition can directly replace Prometheus storage; the cluster edition splits reads, writes, and storage across vmstorage / vminsert / vmselect, giving it stronger horizontal scaling.微信公众号MiddlewareContainersVictoriaMetricsPrometheusMonitoringOperationshttps://www.chenshaowen.com/en/blog/container-deploy-clickhouse.htmlWed, 11 Feb 2026 00:00:00 +0000Wed, 11 Feb 2026 00:00:00 +0000https://www.chenshaowen.com/en/blog/container-deploy-clickhouse.html1. ClickHouse Single Node 1.1 Configure Environment Variables 1 2 3 4 5 6 7 8 export CONTAINER_CLI=nerdctl export IMAGE=clickhouse/clickhouse-server:24 export CLICKHOUSE_INSTANCE_NAME=clickhouse export CH_DATA=/data/ops/clickhouse/$CLICKHOUSE_INSTANCE_NAME export CLICKHOUSE_PORT=9000 export CLICKHOUSE_PROMETHEUS_PORT=9363 export CLICKHOUSE_USER=default export CLICKHOUSE_PASSWORD=xxxxxx 1.2 Generate the Configuration Files 1 2 3 4 5 6 7 8 9 10 11 12 13微信公众号MiddlewareContainersClickHousePrometheusMonitoringOperationshttps://www.chenshaowen.com/en/blog/container-deploy-redis.htmlWed, 11 Feb 2026 00:00:00 +0000Wed, 11 Feb 2026 00:00:00 +0000https://www.chenshaowen.com/en/blog/container-deploy-redis.html1. Redis Single Instance Configure environment variables 1 2 3 4 5 6 7 8 9 export CONTAINER_CLI=nerdctl export IMAGE=redis:7 export REDIS_INSTANCE_NAME=redis-instance export REDIS_PASSWORD=xxxxxx export REDIS_PORT=6379 export REDIS_DATA=/data/ops/redis/$REDIS_INSTANCE_NAME mkdir -p $REDIS_DATA/data $REDIS_DATA/log Start the service 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 $CONTAINER_CLI run -d \ --name $REDIS_INSTANCE_NAME \ --restart always \ --security-opt apparmor=unconfined \ --security-opt seccomp=unconfined \ --network host \ --ulimit memlock=-1 \ --ulimit stack=67108864 \ --ulimit nofile=1048576:1048576 \ --memory-swappiness=0 \ -v $REDIS_DATA/data:/data \ -v $REDIS_DATA/log:/var/log/redis \ $IMAGE redis-server --port $REDIS_PORT --requirepass $REDIS_PASSWORD \ --maxmemory 64gb \ --maxmemory-policy noeviction \ --appendonly yes \ --appendfsync everysec Test the connection A PONG response means the connection succeeded微信公众号MiddlewareContainersRedisOperationshttps://www.chenshaowen.com/en/blog/container-deploy-multi-node-foundationdb-and-maintenance.htmlTue, 10 Feb 2026 00:00:00 +0000Tue, 10 Feb 2026 00:00:00 +0000https://www.chenshaowen.com/en/blog/container-deploy-multi-node-foundationdb-and-maintenance.html1. Generate a Cluster ID 1 cat /dev/urandom | tr -dc 'a-zA-Z0-9' | fold -w 16 | head -n 1 The examples below use CLUSTER_ID=fKbIga9RHP79OIx1. 2. On the First Node Configure environment variables 1 2 3 4 5 6 7 8 9 10 export CONTAINER_CLI=nerdctl export IMAGE=foundationdb/foundationdb:7.1.26 export CLUSTER_ID=fKbIga9RHP79OIx1 export FDB_INSTANCE_NAME=fdb_server export FDB_CLUSTER_FIRST_IP=$(hostname -I | awk '{print $1}') export FDB_PUBLIC_IP=$(hostname -I | awk '{print $1}') export FDB_PORT=4500 export FDB_DIR=/data/ops/fdb/$FDB_INSTANCE_NAME Clean up old data 1 2 3 $CONTAINER_CLI rm -f $FDB_INSTANCE_NAME mv $FDB_DIR $FDB_DIR.微信公众号FoundationDBDatabaseMiddlewareStorageContainers3FSOperationshttps://www.chenshaowen.com/en/blog/message-queue-about-rabbitmq.htmlSat, 23 Dec 2017 20:51:22 +0000Sat, 23 Dec 2017 20:51:22 +0000https://www.chenshaowen.com/en/blog/message-queue-about-rabbitmq.html1. Where a Message Queue Fits 1.1 Asynchronous Processing Use case: after a user registers, a registration email and a registration SMS must be sent. With the synchronous approach, system performance (concurrency, throughput, response time) hits a bottleneck. 1.2 Application Decoupling Use case: after a user places an order, the微信公众号RabbitMQMiddlewareCeleryPythonMessageQueuehttps://www.chenshaowen.com/en/blog/error-code-design-and-unified-processing-in-django.htmlMon, 30 Oct 2017 21:50:33 +0000Mon, 30 Oct 2017 21:50:33 +0000https://www.chenshaowen.com/en/blog/error-code-design-and-unified-processing-in-django.htmlI currently use Django for SaaS development, and I develop and maintain several SaaS applications at the same time. Many SaaS applications have agreed-upon error codes, some used to handle login state, others to mark business logic status. For a feature that is so strongly shared across projects, it is微信公众号DjangoPythonError CodesMiddlewareR&DLearningAPIBest Practices