yuan

proto decoded message size 1195986768 is larger than maximum supported size 4194304

1

There are many errors in the M3Coordinator log. What do these errors mean? the defaultMaxMessageSize is 4MB, how do I modify it. ``` {"level":"error","ts":1632119119.2488701,"msg":"could not read message from consumer","error":"unexpected EOF"}...

M3DB unable to mark namespace as ready without force params.

3

M3DB unable to mark namespace as ready without force params. ``` curl -X POST http://localhost:7201/api/v1/services/m3db/namespace/ready -d '{ "name": "default" } {"status":"error","error":"could not find db session for namespace: default"} ``` ```...

write metrics using http json API only once，But m3coordinator/m3query can look up five minutes of metrics。

9

It looks like an extra 5 minutes metrics are written automatically。Expect to see only one metric at write time. ``` curl -X POST http://localhost:7201/api/v1/json/write -d '{ "tags": { "__name__": "third_avenue",...

io.lettuce.core.RedisCommandTimeoutException: Command timed out after 1 second(s)

3

As long as the command is not executed for a period of time, there will be a timeout problem when the command is executed again. I'm pretty sure that Lettuce...

ERR max number of clients reached

2

线上运行一段时候后，客户端开始报错，Tendis达到连接上限，同时tendis内部日志也开始出现ERR max number of clients reached异常 而业务的实际连接数其实仅仅只有十几个，但是内部通信端口的连接数却有9000+，有的高达2w+。  问题1：Tendis为客户端连接数做限制时，是否包含内部通信的连接数？ 问题2：对于内部通信端口，这些连接的作用是什么？如何减少这些连接数？ 版本：2.3.4-rocksdb-v5.13.4 集群规模：三主三从 配置： ``` bind 0.0.0.0 port 51001 daemon on loglevel notice logdir /tendis/home/log dumpdir /tendis/home/dump dir /tendis/home/db rocks.wal_dir /tendis/home/wal...

为什么master与slave的binlog文件数会差很多？

3

版本：2.3.1 三主三从，上面是master的binlog，下面是slave的

如果一个槽位的主从都挂了，那么整个集群将不可读写，包括其他槽位

6

三主三从，槽位分成三份，先kill其中一个slave，再kill掉对应master 此时整个集群被标记为down，无论客户端访问的key在哪个槽位，都会提示`(error) CLUSTERDOWN The cluster is down` 这种状态不可恢复，除非人工介入。 另外，如果master挂了，看代码slave是以上次同步时间来决定是否有数据落后，进而决定是否可以晋升为主。一旦master退出缓慢（比如：刷盘阻塞），data_age将会比较大，导致slave无法晋升为主，最可怕的是集群会被标记为down，客户端表现仍然为`(error) CLUSTERDOWN The cluster is down`，整个集群灾难级别的不可用。 ``` uint64_t syncTime = replMgr->getLastSyncTime(); data_age = msSinceEpoch() - syncTime; //... mstime_t limitTime = ((mstime_t)gBinlogHeartbeatSecs *...

Introduce interface to calculate remaining validity and renewal for a Lease using `SecretLeaseContainer`

3

expiryThreshold is shared by minRenewal, It is difficult to assess how much minRenewal should be set ``` renewed.getLeaseDuration().getSeconds() < this.minRenewal.getSeconds() ``` https://github.com/spring-projects/spring-vault/blob/afdbdd67f1018634a536fa4e0f60a52d37b1860c/spring-vault-core/src/main/java/org/springframework/vault/core/lease/SecretLeaseContainer.java#L594C5-L594C5