You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@flink.apache.org by "uharaqo (Jira)" <ji...@apache.org> on 2022/06/28 18:42:00 UTC
[jira] [Comment Edited] (FLINK-28206) EOFException on Checkpoint Recovery
|  |
[uharaqo](https://issues.apache.org/jira/secure/ViewProfile.jspa?name=uharaqo)
**edited a comment** on [
FLINK-28206](https://issues.apache.org/jira/browse/FLINK-28206)
---|---
|
---
| [Re: EOFException on Checkpoint
Recovery](https://issues.apache.org/jira/browse/FLINK-28206)
---
| Thanks [~yunta] . I was able to can read the metadata file (checkpoint:
58130) which caused the EOF error:
{noformat}
result = {CheckpointMetadata@955} "Checkpoint Metadata"
checkpointId = 58130
operatorStates = {ArrayList@957} size = 11
0 = {OperatorState@960} "OperatorState(operatorID:
5323682bad0857bbc5dc34e8a0198f36, parallelism: 64, maxParallelism: 128,
coordinatorState: (none), sub task states: 64, total size (bytes):
3607732189)"
1 = {OperatorState@961} "OperatorState(operatorID:
b0d7d72b41ddffa0e746c98bfade0740, parallelism: 64, maxParallelism: 128,
coordinatorState: (none), sub task states: 64, total size (bytes): 11901811)"
2 = {OperatorState@962} "OperatorState(operatorID:
799506d4ab267dabfd7e5e151c32b771, parallelism: 64, maxParallelism: 128,
coordinatorState: (none), sub task states: 0, total size (bytes): 0)"
3 = {OperatorState@963} "OperatorState(operatorID:
bc764cd8ddf7a0cff126f51c16239658, parallelism: 64, maxParallelism: 128,
coordinatorState: 528 bytes, sub task states: 64, total size (bytes): 16400)"
4 = {OperatorState@964} "OperatorState(operatorID:
6ac701675b3507fc60576d5eb1d2a02a, parallelism: 64, maxParallelism: 128,
coordinatorState: (none), sub task states: 0, total size (bytes): 0)"
5 = {OperatorState@965} "OperatorState(operatorID:
4f12c5e521389e5e6c62d38d49148502, parallelism: 64, maxParallelism: 128,
coordinatorState: (none), sub task states: 64, total size (bytes): 0)"
6 = {OperatorState@966} "OperatorState(operatorID:
2963852293169ba90d9d1e7d6308db5c, parallelism: 64, maxParallelism: 128,
coordinatorState: 144 bytes, sub task states: 64, total size (bytes): 15296)"
7 = {OperatorState@967} "OperatorState(operatorID:
30ddfdac31a80b264a11886ca4e04ea8, parallelism: 64, maxParallelism: 128,
coordinatorState: (none), sub task states: 64, total size (bytes): 15868786)"
8 = {OperatorState@968} "OperatorState(operatorID:
ee7af715df0b4d5489c7e945244bab03, parallelism: 64, maxParallelism: 128,
coordinatorState: (none), sub task states: 64, total size (bytes): 646402101)"
9 = {OperatorState@969} "OperatorState(operatorID:
feca28aff5a3958840bee985ee7de4d3, parallelism: 64, maxParallelism: 128,
coordinatorState: 128 bytes, sub task states: 64, total size (bytes): 15264)"
10 = {OperatorState@970} "OperatorState(operatorID:
99f06a2f6abbb4a5ff119455ad01a44b, parallelism: 64, maxParallelism: 128,
coordinatorState: (none), sub task states: 64, total size (bytes): 0)"
masterStates = {Collections$EmptyList@958} size = 0
{noformat}
We use Kubernetes and flink-s3-fs-presto. The network was stable at the time
and we were able to restart the job by ignoring the savepoint and checkpoint.
Do you think the EOF error was caused by the Presto plugin?
---
| | [ 
](https://issues.apache.org/jira/browse/FLINK-28206#add-comment "Add Comment")
| [Add Comment](https://issues.apache.org/jira/browse/FLINK-28206#add-comment
"Add Comment")
---|---
| This message was sent by Atlassian Jira (v8.20.10#820010-sha1:ace47f9) | |
---