分享一个大多人不知道的上海医保知识点, 个人账户缴得越多越吃亏

本市职工医保门急诊就医采用“三段式”保障模式，即：①账户段，先使用每年 7 月 1 日计入的个人账户当年资金支付门急诊费用; ②自负段(即门诊起付线 500)，账户资金用完后，由个人现金支付门急诊费用; ③共付段，超出自负段以上门急诊费用，统筹基金按比例报销

127061

第二个是将来时态，还没有行政令。
Trump Says the U.S. Will Institute $100,000 Fee for Skilled Worker Visas

这个差别可大了，Trump Always Chickens Out (TACO)

不要 rewrite query parameter

nginx.ingress.kubernetes.io/configuration-snippet: |
if ($request_uri ~ "/dhq/([^\?]*)") {
set $path1 $1;
rewrite ^/dhq/ /$path1 break;
}

这个不是 ingress 的问题！
你只是恰巧用了 nginx ，解决方法
https://my.f5.com/manage/s/article/K000148648
试试在 configuration-snippet 再加一个
rewrite ^ $request_uri;

删掉 PV 后重建，记得 spec.claimRef 保持和之前的一样，PVC 会重新 Bound

while read share;do
echo "check $share"
kubectl get pv $share -o yaml > "$share-origin.yaml"
ns=$(yq '.spec.claimRef.namespace' "$share-origin.yaml")
pvc=$(yq '.spec.claimRef.name' "$share-origin.yaml")
echo "$ns $pvc"
if [ ! -f "all-pods-$ns.yaml" ] ;then
kubectl get pods -n $ns -o yaml > "all-pods-$ns.yaml"
fi
podname=$(yq ".items[]|select(.spec.volumes[].persistentVolumeClaim.claimName==\"$pvc\")|.metadata.name" "all-pods-$ns.yaml")
echo "pods $podname will be deleted"
# patch pv
kubectl patch pv $share -p '{"spec":{"persistentVolumeReclaimPolicy":"Retain"}, "metadata":{"finalizers":null }}' --type=merge
kubectl delete pv $share --wait=false
sleep 5
kubectl delete pod -n $ns $podname
sleep 10
kubectl get pv $share || echo deleted
# recreate PV here
sed -e "REPLACE-HERE" "$share-origin.yaml" >"$share-new.yaml"
kubectl create -f "$share-new.yaml"
done<pv-shares.txt

ctr 不走 containerd 的 plugin 配置，plugins."io.containerd.grpc.v1.cri".registry 对 ctr 无用

The ctr utility is a command-line interface for directly interacting with the containerd daemon. It uses containerd's native API, not the CRI plugin's API. Therefore, when you use ctr image pull, it does not consult the registry mirror configurations set up for the CRI plugin.

你如果认为这里应该是或的关系，那改成|就行了

(?x) 这个表示 Free-Spacing ， 后面的空格和换行会 ignore
https://www.regular-expressions.info/freespacing.html

用了新的 feature, range over functions

https://go.dev/blog/range-functions
As of Go 1.23 it now supports ranging over functions that take a single argument. The single argument must itself be a function that takes zero to two arguments and returns a bool; by convention, we call it the yield function.

```
func(yield func() bool)

func(yield func(V) bool)

func(yield func(K, V) bool)
```

胃痉挛疼的厉害的时候一滴水都喝不下，只能打阿托品舒缓平滑肌

> 需要疯狂重启，才有机会点亮
这个试试关掉 UEFI 里的 fast boot 试试，fast boot 打开的时候关机不会重置硬件的状态

如果运维有能力和时间的当然自建。举个例子，k8s 升级后发现新的版本有问题，想回退到上一个版本，很多云上都不支持，自建的话就灵活许多

# ptcpdump is a tcpdump-compatible packet analyzer powered by eBPF, automatically annotating packets with process/container/pod metadata when detectable.
https://github.com/mozillazg/ptcpdump
# audit
auditctl -a exit,always -F arch=b64 -S connect -k "connect-audit"

好像还挺普遍的？
https://www.reddit.com/r/ChatGPT/comments/1k11yw5/after_5_years_of_jaw_clicking_tmj_chatgpt_cured/

11 楼老哥都给你答案了，再给你个详细的
git filter-branch --tree-filter 'cp -f dir/a1 dir/a3' --tag-name-filter cat --prune-empty -- --all

kafka JVM 的 thread stack, heap dump 有吗？
dockerd 的 stack, trace 有吗？
没有的话只能靠猜

我 copy/paste 一个别人的答案
https://groups.google.com/g/golang-nuts/c/SNRW-f1F9aM
Copyright Ian Lance Taylor

HeapInUse tells you the amount of system memory allocated to hold Go
objects, which includes space allocated to hold objects that do not
yet exist or that have been released by the garbage collector. pprof
tells you about live memory that is accessible by your program. With
the default GOGC value of 100, in a program in steady state, it's
normal for HeapInUse to be around twice as much as pprof memory, and
that is what you are seeing.