tiflash表同步进度一直为0

[Hacker_56ngcmM2]

[Lucien]

通过 tiup ctl 查一下这几个 region 的状态，现在看一直在找 region 的 leader，但是 leader 状态不可用。

region {1485273,1101,38} find error: peer is not leader for region 1485273, leader may Some(id: 2703001 store_id: 4)
region {3429817,1101,38} find error: peer is not leader for region 3429817, leader may Some(id: 3429819 store_id: 3)
region {3429817,1101,38} find error: peer is not leader for region 3429817, leader may Some(id: 3429820 store_id: 4)

[Hacker_56ngcmM2]

» region 1485273
{
“id”: 1485273,
“start_key”: “6D44427300000000FF00FA000000000000FF006844423A323934FF3700FE0000000000FA”,
“end_key”: “6D44444C4A6F6248FF69FF73746F727900FF0000FC0000000000FF0000680000000000FF008D08FF00000000FF00000000F7000000FC”,
“epoch”: {
“conf_ver”: 1101,
“version”: 38
},
“peers”: [
{
“id”: 2695201,
“store_id”: 1
},
{
“id”: 2701249,
“store_id”: 3
},
{
“id”: 2703001,
“store_id”: 4
},
{
“id”: 2773936,
“store_id”: 5
},
{
“id”: 2782909,
“store_id”: 2
}
],
“leader”: {
“id”: 2782909,
“store_id”: 2
},
“written_bytes”: 1593,
“read_bytes”: 0,
“written_keys”: 1,
“read_keys”: 0,
“approximate_size”: 92,
“approximate_keys”: 81920
}

» region 1101
null

» region 38
null

» region 3429817
{
“id”: 3429817,
“start_key”: “”,
“end_key”: “6D44427300000000FF00FA000000000000FF006844423A323934FF3700FE0000000000FA”,
“epoch”: {
“conf_ver”: 1101,
“version”: 38
},
“peers”: [
{
“id”: 3429818,
“store_id”: 1
},
{
“id”: 3429819,
“store_id”: 3
},
{
“id”: 3429820,
“store_id”: 4
},
{
“id”: 3429821,
“store_id”: 5
},
{
“id”: 3429822,
“store_id”: 2
}
],
“leader”: {
“id”: 3429821,
“store_id”: 5
},
“written_bytes”: 0,
“read_bytes”: 0,
“written_keys”: 0,
“read_keys”: 0,
“approximate_size”: 41,
“approximate_keys”: 4540
}

» region 2703001
null

» region 3429819
null

» region 3429820
null

[Lucien]

看 region 状态是正常，那有现在有新的报错吗？

[Hacker_56ngcmM2]

别的新的报错没有，就是一直有这种报错

[Lucien]

同步的表是有哪些 DML 操作吗 ？看起来数据一直在 ignore ，有批量定时 update 的操作吗？ 量大吗？可以尝试增加 replicate 的并发速度，看看行不行。另外检查一下 TiKV 的性能带宽。