chihe
|
bf9fdff28d
|
feat(master): save previous decommission error msg when decommissioning another replica
Signed-off-by: chihe <chihe@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
chihe
|
dd9b9be7f5
|
fix(datanode): remove dp 0 from disk error set when recovering disk err
Signed-off-by: chihe <chihe@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
chihe
|
a2724225ec
|
fix(master): enable disk when excuting datnode cancel decommission
Signed-off-by: chihe <chihe@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
chihe
|
44f9fce264
|
fix(master): when rolling back special replica dp.
only need to delete decommission src if new replica is recovered
Signed-off-by: chihe <chihe@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
baihailong
|
e07aeac619
|
fix(fuse): if vol name not match regexp when mount, return failed immediately.
Signed-off-by: baihailong <baihailong@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
chihe
|
6cc809434a
|
fix(master): initialize reportTime for datanode when leader changed
Signed-off-by: chihe <chihe@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
shuqiang-zheng
|
2a9d99bdf9
|
fix(datanode): remove the gc operation to reduce the time required to create a dp.
Signed-off-by: shuqiang-zheng <zhengshuqiang@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
chihe
|
eb1daccedc
|
fix(master): show ignore decommission dp when displaying decommission progress for dataNode
Signed-off-by: chihe <chihe@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
chihe
|
a8d8c96072
|
fix(master): do not delete decommission disk from list when recover task is submitted
Signed-off-by: chihe <chihe@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
chihe
|
d479a79f73
|
fix(master): optimize the calculation of datanode decommission progress
Signed-off-by: chihe <chihe@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
chihe
|
0ffe25c31d
|
fix(datanode): reload dp by parallelism during recovering bad disk
Signed-off-by: chihe <chihe@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
chihe
|
67f5db91ce
|
fix(datanode): only print root path in log
Signed-off-by: chihe <chihe@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
chihe
|
7d05d431e6
|
fix(datanode): optimize the process of constructing heartbeat packets.
Signed-off-by: chihe <chihe@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
chihe
|
8ad898a1e4
|
fix(master): modify logic of updateDecommissionStatus for datanode
Signed-off-by: chihe <chihe@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
chihe
|
39d2baa37e
|
feat(master): add requestID for sending admin task
Signed-off-by: chihe <chihe@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
Victor1319
|
69f6960be9
|
fix(metanode): fix the issue that can't clear data when delete inode.
Signed-off-by: Victor1319 <zengxuewei@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
baihailong
|
8f48f3ac85
|
fix(client): if vol not exists, mount return failed immediately.
Signed-off-by: baihailong <baihailong@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
chihe
|
f02d60be3d
|
fix(master): remove disk from decommission list if it is recoverd
Signed-off-by: chihe <chihe@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
chihe
|
bb0f337f9e
|
fix(master): do not decommission disk with cancel status
Signed-off-by: chihe <chihe@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
chihe
|
7a29ac6ba1
|
fix(datanode): reset diskErrorCnt when recovering bad disk
Signed-off-by: chihe <chihe@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
baihailong
|
21463372eb
|
fix(client): when volume is deleted, cfs-client need to exit.
Signed-off-by: baihailong <baihailong@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
chihe
|
bc2867049f
|
fix(datanode): fix panic raised by function of ExistDir
Signed-off-by: chihe <chihe@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
Victor1319
|
3d2137e545
|
refactor(client): reduce buffer chan size to reduce memory when client idle.
Signed-off-by: Victor1319 <zengxuewei@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
Victor1319
|
ac5cde2a05
|
fix(client): fix the issuse of continuous memory growth on client side.
Signed-off-by: Victor1319 <zengxuewei@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
chihe
|
43d8903d10
|
fix(master): skip auto add replica when dp lost leader
Signed-off-by: chihe <chihe@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
chihe
|
292f2abd45
|
feat(master): add api for recovering bad disk
Signed-off-by: chihe <chihe@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
baihailong
|
75f8f1e07a
|
fix(sdk): access nil pointer leadto cfs-client exit.
Signed-off-by: baihailong <baihailong@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
Victor1319
|
0e7ab29565
|
refactor(client): add config to control stream op max timeout.
Signed-off-by: Victor1319 <zengxuewei@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
Victor1319
|
b0faae7248
|
refactor(client): amplify retry interval once overwrite failed to avoid disk io busy.
Signed-off-by: Victor1319 <zengxuewei@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
chihe
|
c297ded67e
|
fix(master): mark disk as disabled when decommissioning disk
Signed-off-by: chihe <chihe@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
Victor1319
|
857d71602a
|
refactor(data): use tryRun to limit delete io.
Signed-off-by: Victor1319 <zengxuewei@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
Victor1319
|
c11b55c18b
|
refactor(client): refactor random write retry strategy.
Signed-off-by: Victor1319 <zengxuewei@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
NaturalSelect
|
d9ee4e5130
|
feat(data): limit memory used
Signed-off-by: NaturalSelect <huangzhibin1@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
NaturalSelect
|
313c9b0448
|
fix(master): cluster info show readable time
Signed-off-by: NaturalSelect <huangzhibin1@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
leonrayang
|
1b9b48f846
|
refactor(master): refactor the code of vol struct and cluster struct
:
Signed-off-by: leonrayang <chl696@sina.com>
|
2024-12-26 09:24:22 +08:00 |
|
NaturalSelect
|
c551e5dbc2
|
fix(data): return limited io if failed to delete batch extents
Signed-off-by: NaturalSelect <huangzhibin1@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
NaturalSelect
|
6aa160552c
|
feat(master): persist datanode bad disk
Signed-off-by: NaturalSelect <huangzhibin1@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
chihe
|
f36c40c07f
|
fix(master):if all replica is unvaliable, mark dp as failed
Signed-off-by: chihe <chihe@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
chihe
|
942adaed7d
|
feat(datanode): add debug log for heartbeat
Signed-off-by: chihe <chihe@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
NaturalSelect
|
5365949f7c
|
feat(master): support config decommission settings
Signed-off-by: NaturalSelect <huangzhibin1@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
NaturalSelect
|
8386685205
|
feat(master): parallelism check dp meta
Signed-off-by: NaturalSelect <huangzhibin1@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
true1064
|
8401562ef6
|
fix(master): only choose from the specific zone in function canWriteForNode
Signed-off-by: true1064 <tangjingyu@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
chihe
|
54402d030a
|
fix(master): clean decommission list when leader change
Signed-off-by: chihe <chihe@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
chihe
|
442714ad20
|
fix(master): wait for setting RestoreReplica of dp during MarkDecommissionStatus
Signed-off-by: chihe <chihe@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
chihe
|
145e53d2f6
|
fix(datanode): rename root dir of dp to dir with backup prefix.
when execute decommisison with raftForce
Signed-off-by: chihe <chihe@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
chihe
|
04ac20e2cb
|
fix(master): calculating the decommission progress of the disk includes ignoring dps.
Signed-off-by: chihe <chihe@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
chihe
|
a650776bbf
|
fix(master): execute recoverReplicaMeta when replica is IOError
Signed-off-by: chihe <chihe@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
baihailong
|
a98a67358d
|
fix(libsdk): add IsDir and IsRegular for java.
Signed-off-by: baihailong <baihailong@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
baihailong
|
92edf40c3e
|
fix(fuse): MountOption add parameter DisableMountSubtype
|
2024-12-26 09:24:22 +08:00 |
|
chihe
|
cab9ec44f3
|
fix(master): dp is decommissioned by raftForce, disk manager check recovery for new replica
close:#21938887
Signed-off-by: chihe <chihe@oppo.com>
|
2024-12-26 09:24:22 +08:00 |
|
