Lily2025 comments

Results 106 comments of


                                            Lily2025

ticdc restart and changefeed lag reached more than 10min when inject network partition betweent pdleader and pdfollowers

inject network partition betweent pdleader and pdfollowers ![image](https://github.com/pingcap/tiflow/assets/84712107/a252f930-4194-4f31-af34-ffce16931001) two ticdc crash ![image](https://github.com/pingcap/tiflow/assets/84712107/ede86b45-9ad1-492b-a9d6-e56be02a1e99) ![image](https://github.com/pingcap/tiflow/assets/84712107/da5733ce-e3d4-49a0-b127-0e80d89627d2)

ticdc restart and changefeed lag reached more than 10min when inject network partition betweent pdleader and pdfollowers

inject network partition between ticdc owner and all other pods，ticdc restart chaos start ~ chaos end：2024/02/28 19:05:36 ~ 2024/02/28 19:08:36 ![img_v3_028h_eceea3ca-f765-4c15-b4b3-1b92b7220d4g](https://github.com/pingcap/tiflow/assets/84712107/4f3f18b3-2005-458b-93d3-0cceef86241c) ticdc logs： [2024/02/28 19:08:37.410 +08:00] [ERROR] [tso_dispatcher.go:562] ["[tso] update...

ticdc lag reached more than 12min when run ha_ticdc(owner)_to_ticdcProcessor(all)_network_partition

/remove-area dm /area ticdc

[Dynamic Regions] tracking capacity reporting exception of pd、tikv cpu is not balance due to schedule slowly and region number is abnormally too much when tpcc prepare 2w warehouse and run workload

/type bug /severity critical /assign bufferflies

add index failed with error “Error 1105 (HY000): request pd http api failed with status: '503 Service Unavailable” when kill pd leader

/type bug /severity major /assign rleungx

Lily2025

ticdc restart and changefeed lag reached more than 10min when inject network partition betweent pdleader and pdfollowers

ticdc restart and changefeed lag reached more than 10min when inject network partition betweent pdleader and pdfollowers

ticdc lag reached more than 12min when run ha_ticdc(owner)_to_ticdcProcessor(all)_network_partition

[Dynamic Regions] tracking capacity reporting exception of pd、tikv cpu is not balance due to schedule slowly and region number is abnormally too much when tpcc prepare 2w warehouse and run workload

pd panic when inject pd leader io delay 1s last for 5mins

QPS fell to zero for a few minutes after fault recover from minority tikv network-loss due to full of "Async Apply CPU"

add index is always running after injecting pd leader io delay 1s last for 1mins and then recover

ticdc restart and changefeed lag reached more than 10min when inject network partition betweent pdleader and pdfollowers

ticdc restart and changefeed lag reached more than 10min when inject network partition betweent pdleader and pdfollowers

add index failed with error “Error 1105 (HY000): request pd http api failed with status: '503 Service Unavailable” when kill pd leader