Windows下docker compose up -d遇到错误

[huangxf14]

✘ Container coze-milvus Error 36.1s
✔ Container coze-server Created 0.0s
✘ Container coze-elasticsearch Error 73.3s
dependency failed to start: container coze-milvus is unhealthy

根据issue里把setup.sh下的CRLF改了。开了梯子全局模式。

docker log -f coze-elasticsearch：
......
Checking Elasticsearch availability... (attempt 60)
Elasticsearch not available, retrying in 1 seconds...
-e 🔍 Checking smartcn plugin status...
-e ❌ smartcn plugin not loaded correctly, please ensure the plugin is installed and Elasticsearch is restarted
Elasticsearch initialization completed successfully!

[hi-pender]

这个是哪个版本的镜像？ 最新版的镜像会使用 coze-studio/docker/volumes/elasticsearch 目录下事先下载好的 smartcn 插件。 请更新使用 coze-studio 的代码仓库最新版 和 coze-server 镜像的最新版哈

或者是有个是 elasticsearch 第一次启动时，smartcn 未完成安装就退出了，这种情况需要删除 elasticsearch 数据，重新启动镜像

[hzl-123]

我通过centos部署拉取的最新代码也是启动不了，报这个错

[hzl-123]

这个是哪个版本的镜像？ 最新版的镜像会使用 coze-studio/docker/volumes/elasticsearch 目录下事先下载好的 smartcn 插件。 请更新使用 coze-studio 的代码仓库最新版 和 coze-server 镜像的最新版哈

或者是有个是 elasticsearch 第一次启动时，smartcn 未完成安装就退出了，这种情况需要删除 elasticsearch 数据，重新启动镜像

怎么删除elasticsearch 数据呢

[hi-pender]

docker/data 目录，是各基础组件的数据目录。 里面有个 docker/data/bitnami/elasticsearch 目录

[hzl-123]

已解决，通过更改docker-compose.yml配置文件里的etcd挂载配置，删掉挂载的 etcd.conf.yml

[hi-pender]

已解决，通过更改docker-compose.yml配置文件里的etcd挂载配置，删掉挂载的 etcd.conf.yml

有了解到是什么原因么？ 挂在一个配置文件，为啥会导致出现 CentOS 无法启动的兼容性问题

[jachinpy]

查看对应 elasticsearch 的healthy ， 如果error， 尝试增大等待时间。 猜想是由于不同内核版本的文件系统对于挂载文件的读取效率有所差异。

start_period: 60s

[GXKIM]

✘ Container coze-milvus Error 36.1s ✔ Container coze-server Created 0.0s ✘ Container coze-elasticsearch Error 73.3s dependency failed to start: container coze-milvus is unhealthy

根据issue里把setup.sh下的CRLF改了。开了梯子全局模式。

docker log -f coze-elasticsearch： ...... Checking Elasticsearch availability... (attempt 60) Elasticsearch not available, retrying in 1 seconds... -e 🔍 Checking smartcn plugin status... -e ❌ smartcn plugin not loaded correctly, please ensure the plugin is installed and Elasticsearch is restarted Elasticsearch initialization completed successfully!

我是重启几次就可以了，不再提示

[immengzi]

重置 Docker 网络解决问题。
[root@openEuler docker]# docker-compose up -d
[+] Running 9/9
✔ Container coze-nsqlookupd Healthy 0.6s
✔ Container coze-nsqd Running 0.0s
✔ Container coze-nsqadmin Running 0.0s
✔ Container coze-etcd Healthy 0.6s
✔ Container coze-elasticsearch Healthy 1.1s
✔ Container coze-mysql Healthy 1.1s
✔ Container coze-minio Healthy 1.1s
✘ Container coze-milvus Error 11.0s
✔ Container coze-redis Healthy 1.3s
dependency failed to start: container coze-milvus is unhealthy
[root@openEuler docker]# systemctl stop firewalld
[root@openEuler docker]# systemctl stop docker
[root@openEuler docker]# iptables -F
[root@openEuler docker]# systemctl start docker
[root@openEuler docker]# docker-compose up -d
[+] Running 11/11
✔ Container coze-etcd Healthy 0.6s
✔ Container coze-mysql Healthy 19.6s
✔ Container coze-minio Healthy 19.6s
✔ Container coze-milvus Healthy 19.6s
✔ Container coze-redis Healthy 19.6s
✔ Container coze-nsqlookupd Healthy 0.6s
✔ Container coze-nsqd Running 0.0s
✔ Container coze-elasticsearch Healthy 19.6s
✔ Container coze-nsqadmin Running 0.0s
✔ Container coze-server Started 2.7s
✔ Container coze-web Started 9.1s
[root@openEuler docker]#