View Issue Details
ID | Project | Category | View Status | Date Submitted | Last Update |
---|---|---|---|---|---|
0005602 | SymmetricDS | Bug | public | 2022-11-28 19:33 | 2022-11-30 19:40 |
Reporter | elong | Assigned To | elong | ||
Priority | normal | ||||
Status | closed | Resolution | fixed | ||
Product Version | 3.13.0 | ||||
Target Version | 3.13.9 | Fixed in Version | 3.13.9 | ||
Summary | 0005602: Parse exception of batch should remove it from staging | ||||
Description | When an update is read from a batch CSV, if the elements are less than the column count, it goes into a loop with a ParseException. It should remove the batch from staging with a ProtocolException and let the source node know to do the same. ERROR [client-1] [ManageIncomingBatchListener] [client-1-dataloader-8] Failed to load batch server-26 java.lang.IllegalStateException: Invalid state while parsing csv data. The number of columns (2) reported for table 'MYTEST' doesn't match up with the token count (2) data: {update,1} at org.jumpmind.symmetric.io.data.reader.ProtocolDataReader.readNext(ProtocolDataReader.java:194) at org.jumpmind.symmetric.io.data.reader.ProtocolDataReader.nextData(ProtocolDataReader.java:367) at org.jumpmind.symmetric.io.data.DataProcessor.forEachDataInTable(DataProcessor.java:181) at org.jumpmind.symmetric.io.data.DataProcessor.forEachTableInBatch(DataProcessor.java:160) at org.jumpmind.symmetric.io.data.DataProcessor.process(DataProcessor.java:108) at org.jumpmind.symmetric.service.impl.DataLoaderService$LoadIntoDatabaseOnArrivalListener$2.call(DataLoaderService.java:1046) at org.jumpmind.symmetric.service.impl.DataLoaderService$LoadIntoDatabaseOnArrivalListener$2.call(DataLoaderService.java:1) at java.util.concurrent.FutureTask.run(FutureTask.java:266) at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624) at java.lang.Thread.run(Thread.java:750) | ||||
Steps To Reproduce | - create table mytest (id integer primary key, name varchar(100)); - insert into mytest values (1, 'eric'); - stop client's pull job - update mytest set name = 'poop' where id = 1 - view batch and refresh its sent data - edit the batch in staging to remove some data elements from the update - start client's pull job | ||||
Tags | No tags attached. | ||||
Date Modified | Username | Field | Change |
---|---|---|---|
2022-11-28 19:33 | elong | New Issue | |
2022-11-28 19:33 | elong | Status | new => assigned |
2022-11-28 19:33 | elong | Assigned To | => elong |
2022-11-28 19:37 | elong | Description Updated | View Revisions |
2022-11-28 19:40 | elong | Status | assigned => resolved |
2022-11-28 19:40 | elong | Resolution | open => fixed |
2022-11-28 19:40 | elong | Fixed in Version | => 3.13.9 |
2022-11-28 19:40 | elong | Issue cloned: 0005603 | |
2022-11-28 19:40 | elong | Relationship added | related to 0005603 |
2022-11-28 20:00 | admin | Changeset attached | => SymmetricDS 3.13 0e76685a |
2022-11-30 19:40 | admin | Status | resolved => closed |