[ https://issues.apache.org/jira/browse/FLINK-19486?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17208591#comment-17208591 ]
Aljoscha Krettek commented on FLINK-19486: ------------------------------------------ Setting the RocksDB backend in code will override anything you specify in the command. It seems that you took the savepoint using RocksDB and are then restoring with the {{FsStateBackend}}. That will not work. > expected: class org.apache.flink.runtime.state.KeyGroupsStateHandle, but > found: class org.apache.flink.runtime.state.IncrementalRemoteKeyedStateHandle > ------------------------------------------------------------------------------------------------------------------------------------------------------ > > Key: FLINK-19486 > URL: https://issues.apache.org/jira/browse/FLINK-19486 > Project: Flink > Issue Type: Bug > Components: Runtime / Checkpointing, Runtime / State Backends > Affects Versions: 1.11.1 > Reporter: appleyuchi > Priority: Major > Fix For: 1.11.1 > > Attachments: pom.xml > > > *The reason why I reopen it is that:* > > *I can {color:#de350b}not{color} {color:#de350b}resume from incremental > checkpoint manually{color} when I stop the following program by force(maybe > NOT supported officially?).* > ||My Experience||StateBackEnd||result|| > |Normal Checkpoint|FsStateBackend|succeed| > |Incremental Checkpoint|RocksDBStateBackend|failed| > ||My Other Effort||Content||Result|| > |①|*I {color:#de350b}hava subscribed to mailing list{color}*|*David Anderson > has answered this several days ago. But he did not tell me how I can fix it.* > *I guess he's very busy.* > |②|{color:#de350b}stackoverflow account{color}|it is forbidden to ask > questions. > |③|alibaba {color:#de350b}Flink dingtalk group{color}|no one can answer this > question > |④|I have read > [document|https://ci.apache.org/projects/flink/flink-docs-stable/ops/state/checkpoints.html] > carefully|I can NOT succeed in it| > |⑤|apologise for post it here|how can I make sure this question is a support > instead of a bug or due to my own mistake? > no successful example in Google/Baidu,am I right?| > |⑥|about less than 1000 people visit my [failure > record|https://yuchi.blog.csdn.net/article/details/108896441] about this > experiment.|in vain| > *Please help,thanks~* > > ---------------------------------original posting > content--------------------------------- > *I want to do an experiment of"incremental checkpoint"* > ||information||content|| > |pom.xml|[ is in the attachment] > |code| [https://paste.ubuntu.com/p/DpTyQKq6Vk/] > |error|[https://paste.ubuntu.com/p/49HRYXFzR2/] > > *some of the above error is:* > *Caused by: java.lang.IllegalStateException: Unexpected state handle type, > expected: class org.apache.flink.runtime.state.KeyGroupsStateHandle, but > found: class org.apache.flink.runtime.state.IncrementalRemoteKeyedStateHandle* > > > The steps for incremental experiment are: > ||step||content|| > |①|mvn clean scala:compile compile package| > |②|nc -lk 9999| > |③|flink run -c wordcount_increstate datastream_api-1.0-SNAPSHOT.jar > Job has been submitted with JobID df6d62a43620f258155b8538ef0ddf1b| > |④|input the following conents in nc -lk 9999 > before > error > error > error > error > |⑤|flink run -s > hdfs://Desktop:9000/tmp/flinkck/df6d62a43620f258155b8538ef0ddf1b/chk-22 -c > StateWordCount datastream_api-1.0-SNAPSHOT.jar > Then the above error happens. > Please help,Thanks~! > -- This message was sent by Atlassian Jira (v8.3.4#803005)