关于 docker 23.0.0 和 23.0.1 的一些问题，看看大家有遇到吗？

[maleedingdaa]

我在一些服务器上（ ubuntu20.04 ），使用了这两个版本的 docker ，发现 docker 服务时不时就会崩溃重启，导致我的容器也自己重启了。使的我容器内的一些服务出现了异常。不知道大家有没有遇到，有没有解决办法？或者我只能回退安装 23 版本以前的 docker 了。以下是 docker 重启的日志：

Feb 12 02:20:17 ecs-2383 dockerd[274932]: time="2023-02-12T02:20:17.694078365Z" level=warning msg="[resolver] failed to read from DNS server: 127.0.0.53:53, query: ;yidui>
Feb 12 03:05:30 ecs-2383 dockerd[274932]: panic: runtime error: invalid memory address or nil pointer dereference
Feb 12 03:05:30 ecs-2383 dockerd[274932]: [signal SIGSEGV: segmentation violation code=0x1 addr=0x20 pc=0x55a5805487ff]
Feb 12 03:05:30 ecs-2383 dockerd[274932]: goroutine 11868 [running]:
Feb 12 03:05:30 ecs-2383 dockerd[274932]: github.com/docker/docker/libnetwork.(*resolver).ServeDNS(0xc000f6a540, {0x55a581adf0c0, 0xc000e09900}, 0xc0006565a0)
Feb 12 03:05:30 ecs-2383 dockerd[274932]:         /go/src/github.com/docker/docker/libnetwork/resolver.go:399 +0x6ff
Feb 12 03:05:30 ecs-2383 dockerd[274932]: github.com/docker/docker/vendor/github.com/miekg/dns.(*Server).serveDNS(0xc0000e7b00, {0xc000dba200, 0x25, 0x200}, 0xc000e09900)
Feb 12 03:05:30 ecs-2383 dockerd[274932]:         /go/src/github.com/docker/docker/vendor/github.com/miekg/dns/server.go:651 +0x4e2
Feb 12 03:05:30 ecs-2383 dockerd[274932]: github.com/docker/docker/vendor/github.com/miekg/dns.(*Server).serveUDPPacket(0xc0000e7b00, 0x0?, {0xc000dba200, 0x25, 0x200}, {>
Feb 12 03:05:30 ecs-2383 dockerd[274932]:         /go/src/github.com/docker/docker/vendor/github.com/miekg/dns/server.go:591 +0x185
Feb 12 03:05:30 ecs-2383 dockerd[274932]: created by github.com/docker/docker/vendor/github.com/miekg/dns.(*Server).serveUDP
Feb 12 03:05:30 ecs-2383 dockerd[274932]:         /go/src/github.com/docker/docker/vendor/github.com/miekg/dns/server.go:521 +0x485
Feb 12 03:05:30 ecs-2383 systemd[1]: docker.service: Main process exited, code=exited, status=2/INVALIDARGUMENT
Feb 12 03:05:30 ecs-2383 systemd[1]: docker.service: Failed with result 'exit-code'.
Feb 12 03:05:32 ecs-2383 systemd[1]: docker.service: Scheduled restart job, restart counter is at 4.
Feb 12 03:05:32 ecs-2383 systemd[1]: Stopped Docker Application Container Engine.
Feb 12 03:05:32 ecs-2383 systemd[1]: Starting Docker Application Container Engine...
Feb 12 03:05:32 ecs-2383 dockerd[302526]: time="2023-02-12T03:05:32.643965606Z" level=info msg="Starting up"
Feb 12 03:05:32 ecs-2383 dockerd[302526]: time="2023-02-12T03:05:32.646131806Z" level=info msg="detected 127.0.0.53 nameserver, assuming systemd-resolved, so using resolv>
Feb 12 03:05:32 ecs-2383 dockerd[302526]: time="2023-02-12T03:05:32.650232757Z" level=info msg="[core] [Channel #1] Channel created" module=grpc
Feb 12 03:05:32 ecs-2383 dockerd[302526]: time="2023-02-12T03:05:32.650396954Z" level=info msg="[core] [Channel #1] original dial target is: \"unix:///run/containerd/cont>
Feb 12 03:05:32 ecs-2383 dockerd[302526]: time="2023-02-12T03:05:32.650534499Z" level=info msg="[core] [Channel #1] parsed dial target is: {Scheme:unix Authority: Endpoin>
Feb 12 03:05:32 ecs-2383 dockerd[302526]: time="2023-02-12T03:05:32.650669189Z" level=info msg="[core] [Channel #1] Channel authority set to \"localhost\"" module=grpc
Feb 12 03:05:32 ecs-2383 dockerd[302526]: time="2023-02-12T03:05:32.650921914Z" level=info msg="[core] [Channel #1] Resolver state updated: {\n  \"Addresses\": [\n    {\n>
Feb 12 03:05:32 ecs-2383 dockerd[302526]: time="2023-02-12T03:05:32.652209595Z" level=info msg="[core] [Channel #1] Channel switches to new LB policy \"pick_first\"" modu>
Feb 12 03:05:32 ecs-2383 dockerd[302526]: time="2023-02-12T03:05:32.652416486Z" level=info msg="[core] [Channel #1 SubChannel #2] Subchannel created" module=grpc
Feb 12 03:05:32 ecs-2383 dockerd[302526]: time="2023-02-12T03:05:32.652572189Z" level=info msg="[core] [Channel #1 SubChannel #2] Subchannel Connectivity change to CONNEC>
Feb 12 03:05:32 ecs-2383 dockerd[302526]: time="2023-02-12T03:05:32.652702899Z" level=info msg="[core] [Channel #1 SubChannel #2] Subchannel picks a new address \"/run/co>
Feb 12 03:05:32 ecs-2383 dockerd[302526]: time="2023-02-12T03:05:32.653339994Z" level=info msg="[core] [Channel #1] Channel Connectivity change to CONNECTING" module=grpc

[pubby]

我是碰到过用基于 ubuntu:22.04 制作的 docker 镜像，而且是那种放一个简单的 go 编译的程序， 在部分 k8s 节点随机崩溃

后来改成 ubuntu:latest 就稳定了

[julyclyde]

可以试试把 daemon 和 container 分离 https://docs.docker.com/config/containers/live-restore/

[maleedingdaa]

@pubby 估计是 docker 自己的问题。查到也有人这个问题，https://forums.docker.com/t/after-upgraded-to-docker-23-0-0-docker-service-crashes-every-night/134593 。

[maleedingdaa]

@julyclyde 非常感谢，刚才试了一下，感觉不错。不知道能不能用在生产环境？