Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Inconsistent replication of data in asynchronous mode of distributed orientdb #3785

Closed
rinkymangal2010 opened this issue Mar 20, 2015 · 6 comments
Assignees
Milestone

Comments

@rinkymangal2010
Copy link

Hi,

I am not able to replicate data to other nodes of distributed cluster with 100% reliabilty.

I have setup a cluster of 3 nodes (node1, node2 and node3).

My default-distributed-db-config.json is like:

{
"autoDeploy": true,
"hotAlignment": false,
"executionMode": "asynchronous",
"readQuorum": 1,
"writeQuorum": 2,
"failureAvailableNodesLessQuorum": false,
"readYourWrites": true,
"clusters": {
"internal": {
},
"index": {
},
"*": {
"servers" : [ "<NEW_NODE>" ]
}
}
}

Steps to reproduce Issue:

  1. I have inserted 200K documents at "node1" with the rate of ~2400 documents/second.
  2. Wait for replication of all 200k documents at node2 and node3.
  3. Deleted data from "node2" and waited for deletion to replicate on all other nodes(node2 and node3).
  4. Again inserted data at "node1" with the same rate.
  5. Again, I was waiting for replication of all 200k documents at node2 and node3.
  6. But no data replicate to "Node2", from where the data were deleted.
  7. While all the data is replicated to node3.

Tried with:
orientdb-2.0.3
orientdb-2.0.6-SNAPSHOT

Regards,
Rinky Mangal

@rinkymangal2010
Copy link
Author

To reproduce the issue easily you may use only 1 document.

I have reproduced the same issue with 1 document.

Please follow below steps:

  1. Insert 1 documents at "node1".
  2. Wait for replication of document at node2 and node3.
  3. Deleted data from "node2" and waited for deletion to replicate on all other nodes(node2 and node3).
  4. Again inserted 1 same document at "node1".
  5. Again, wait for replication document at node2 and node3.
  6. You will found no data replicated to "Node2", from where the data was deleted.
  7. While the data is replicated to node3.

Note:
Also, I was creating index by applying unique constraint on it like:

create index key_usertable_Index on usertable(key) unique;

Please let me know, how can i handle this scenario.

Thamks,
Rinky Mangal

@rinkymangal2010
Copy link
Author

There are two points to note about this issue:

  1. Replication works fine if there is no unique constraint like:
    create class usertable ;
    CREATE PROPERTY usertable.key string;
    CREATE INDEX key_ usertable_Index on usertable(key) NOTUNIQUE;
    insert into usertable (key, field1) values ('key1', 'value1');
  2. Issue can be observed while inserting the data at the same index which were deleted previously and at the node from where the deletion operation were performed.

Thanks,
Rinky Mangal

@lvca
Copy link
Member

lvca commented Mar 25, 2015

Please could you retry against last 2.0.6-SNAPSHOT?

@lvca lvca self-assigned this Mar 25, 2015
@lvca lvca added the bug label Mar 25, 2015
@lvca lvca added this to the 2.0.6 milestone Mar 25, 2015
@rinkymangal2010
Copy link
Author

@lvca Issue still persist.

I have tested for the same with the following snapshot:
orientdb-community-2.0.6-20150325.091043-11-distribution.tar.gz

Link address:
https://oss.sonatype.org/content/repositories/snapshots/com/orientechnologies/orientdb-community/2.0.6-SNAPSHOT/orientdb-community-2.0.6-20150325.091043-11-distribution.tar.gz

@lvca lvca modified the milestones: 2.0.6, 2.0.7 Mar 31, 2015
@lvca lvca removed the bug label Apr 8, 2015
@lvca lvca modified the milestones: 2.0.7, 2.0.8, 2.0.9 Apr 14, 2015
@lvca lvca modified the milestones: 2.0.9, 2.0.10 May 14, 2015
@lvca lvca modified the milestones: 2.0.10, 2.0.11 May 22, 2015
@lvca lvca modified the milestones: 2.0.11, 2.1 GA Jun 19, 2015
@lvca lvca modified the milestones: 2.1 GA, 2.0.11 Jun 19, 2015
@lvca
Copy link
Member

lvca commented Jul 30, 2015

This could be the same as for #2914 just fixed.

@lvca
Copy link
Member

lvca commented Aug 5, 2015

This issue should be the same as for #2914. I'm closing this issue. In case you're able to reproduce it again against last 2.1, please comment/reopen this issue. Thanks.

@lvca lvca closed this as completed Aug 5, 2015
@lvca lvca added the duplicate label Aug 5, 2015
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Development

No branches or pull requests

2 participants