OHazelcastPlugin ... error on reading distributed request: deploy_db

[Frens Jan Rumph]

Dear all,

I am trying to move to a distributed setup of an OrientDB graph database. I'm tinkering around in a local setup:

• version 1.7 from https://oss.sonatype.org/content/repositories/releases/com/orientechnologies/orientdb-community/1.7/orientdb-community-1.7-distribution.zip
• two copies of the distribution in folders 'a' and 'b'
• named the nodes 'orient-a' and 'orient-b' in the com.orientechnologies.orient.server.hazelcast.OHazelcastPlugin section of orientdb-dserver-config.xml
• configured nodes a and b to find each other through the static ip address option in hazelcast.xml (each on their own port 2434 and 2435)
• mentioned 'orient-a' and 'orient-b' in default-distributed-db-config.json

The two instances connect, the node which starts second (I always start 'orient-b' after 'orient-a') drops the GratefulDeadConcerts database in order to replicate from the first node ('orient-a' in this case). However after this I see the following message on the second instance ('orient-b')

2014-05-27 10:57:28:473 WARN [orient-b]->[[orient-a]] requesting deploy of database 'GratefulDeadConcerts' on local server... [OHazelcastPlugin]

On 'orient-a' (the instances started first) I see the messages:

2014-05-27 10:57:26:415 INFO [orient-a]<-[orient-b] added node configuration id=Member [localhost]:2435 name=orient-b, now 2 nodes are configured [OHazelcastPlugin][orient-a]<-[orient-b] error on reading distributed request: deploy_db Distributed storage was not installed for database 'GratefulDeadConcerts'
-> com.orientechnologies.orient.server.hazelcast.OHazelcastPlugin.executeOnLocalNode(OHazelcastPlugin.java:644)
-> com.orientechnologies.orient.server.hazelcast.OHazelcastDistributedDatabase.onMessage(OHazelcastDistributedDatabase.java:477)
-> com.orientechnologies.orient.server.hazelcast.OHazelcastDistributedDatabase$1.run(OHazelcastDistributedDatabase.java:243)
-> java.lang.Thread.run(Thread.java:744)

Anyone have an idea of what I might be doing wrong?

Any help would be much appreciated,

Frens 

[Frens Jan Rumph]

Moreover, when I do insert a record in a database on 'orient-a', I get a message like this in the logs:

2014-05-27 11:26:00:560 WARN [orient-a] overridden quorum (2) for request id=1 from=orient-a task=record_create(#9:-1 v.0) because available nodes (1) are less than quorum, received responses: {orient-a=#9:0 v.1, orient-b=waiting-for-response} ODistributedResponseManager]

What also strikes me as odd is that I am able to connect with both instances via studio. However, connecting with the instances via the console (e.g. with connect remote:localhost:2434 username password or connect remote:localhost:2435 username password) results in the console displaying:

Connecting to remote Server instance [remote:localhost:2435] with user 'username'...

Note that I witness this behaviour for both instances.

Again, any help would be much appreciated ... I'll probably slap myself in the face because of my stupidity ... but I'm not able to find out where I'm going wrong.

Cheers,

Frens

[Artem Orobets]

Hi Frens,

Could you provide full logs from 'orient-a' and 'orient-b'?

Best regards,

Artem Orobets

Orient Technologies

the Company behind OrientDB

--

---
You received this message because you are subscribed to the Google Groups "OrientDB" group.
To unsubscribe from this group and stop receiving emails from it, send an email to orient-databa...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Frens Jan Rumph]

Hi Artem,

Sure, see the files attached.

Br,

Frens Jan

[Frens Jan Rumph]

Anyone have any suggestions?

[Artem Orobets]

Hi Frens,

2014-05-27 12:34 GMT+03:00 Frens Jan Rumph <fren...@frensjan.nl>:

However, connecting with the instances via the console (e.g. with connect remote:localhost:2434 username password or connect remote:localhost:2435 username password) results in the console displaying:

Connecting to remote Server instance [remote:localhost:2435] with user 'username'...

Note that I witness this behaviour for both instances.

You are connecting to the wrong port, try 2424 and 2425. It is a right ports for console.

As for deploy issue, looks like a bug. Could you create a bug report?

[Rumph, Frens Jan]

Hi Artem,

Thanks for the response. Are you sure I'm connecting to the wrong port?

I configured node-a to listen on 2424 and node-b to listen on port 2425 in order to be able to run the two nodes on one single machine / os / ip address.

I'll create a bug report for the deployment issue. Could you recommend a work around?

Cheers,

Frens

--

---
You received this message because you are subscribed to a topic in the Google Groups "OrientDB" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/orient-database/QpZPG4y_KpU/unsubscribe.
To unsubscribe from this group and all its topics, send an email to orient-databa...@googlegroups.com.

[Artem Orobets]

Hi Rumph,

Yes. In console you configured 2424 but in console you specify 2434, that is used for other purposes.

However, connecting with the instances via the console (e.g. with connect remote:localhost:2434 username password or connect remote:localhost:2435 username password) results in the console displaying:

  Connecting to remote Server instance [remote:localhost:2435] with user 'username'...

Best regards,

Artem Orobets

Orient Technologies

the Company behind OrientDB

--

---

You received this message because you are subscribed to the Google Groups "OrientDB" group.

To unsubscribe from this group and stop receiving emails from it, send an email to orient-databa...@googlegroups.com.

[Андрей Логинов]

Hi Artem,

I'm trying OrientDB v.1.7.2 and I'm still facing the deployment issue described by Frens Jan Rumph [1 node has a DB, another node does not, node №2 requests deployment while node №1 fails with an error].
I haven't found a bug report about this issue as yet. Have it been created already?

Meanwhile, is there another option of DB deployment to other nodes?

среда, 4 июня 2014 г., 8:33:10 UTC+3 пользователь Artem Orobets написал:

[Artem Orobets]

Hi Andrey,

It has not been created. As I understand the problem in Frens's case was in his configuration.

Best regards,

Artem Orobets

Orient Technologies

the Company behind OrientDB

--

[Luca Garulli]

Hi,

Please can you try with 1.7.4?

Lvc@

--

[Андрей Логинов]

Hi Artem, Lvc,

Thanks for response. I'm facing this issue with 1.7.4 DB too.
Please check my details:
1) I have 2 nodes running locally, 'Andrew-node-1' and 'Andrew-node-2'.
2) I have configured the database "GeoLocation" on 'Andrew-node-1' and I expect it to be deployed on 'Andrew-node-2'.
3) Hazelcast determines nodes directly by IP. As both nodes are locate on 1 machine, they have different ports. 'Andrew-node-1':2434 and 'Andrew-node-2':2435.

Please check the logs:
Andrew-node-1: http://pastebin.com/ZaUai4wX
Andrew-node-2: http://pastebin.com/xAGhDMmD

I've also attached both json config files.



среда, 25 июня 2014 г., 12:24:05 UTC+3 пользователь Lvc@ написал:

[Андрей Логинов]

Hi all,

I tried this use case with 1.6.4 - works fine, DB is deployed to another node. However json config for 1.6.4 looks a little different from 1.7.*: servers are listed in "partitioning" field.


node2 log:

2014-06-25 15:22:53:770 WARN [Andrew-node-2]<-[Andrew-node-1] installing database GeoLocation in D:\Soft\geo_locations\orientdb-community-1.6.4-node2/databases/GeoLocation... [OHazelcastPlugin]
2014-06-25 15:22:54:363 WARN [Andrew-node-2] installed database GeoLocation in D:\Soft\geo_locations\orientdb-community-1.6.4-node2/databases/GeoLocation, setting it online... [OHazelcastPlugin]
2014-06-25 15:22:54:380 WARN Can not restore 1 WAL master record for storage GeoLocation [OWriteAheadLog]
2014-06-25 15:23:01:893 WARN [Andrew-node-2] database GeoLocation is online [OHazelcastPlugin]
2014-06-25 15:23:01:893 WARN [Andrew-node-2] updated node status to 'ONLINE' [OHazelcastPlugin]

node1 log:

2014-06-25 15:22:24:164 WARN [Andrew-node-1]->[Andrew-node-2] deploying database GeoLocation... [ODeployDatabaseTask]
2014-06-25 15:22:53:408 WARN [Andrew-node-1]->[Andrew-node-2] sending the compressed database GeoLocation over the network, total 10,67Mb [ODeployDatabaseTask]
2014-06-25 15:22:59:507 WARN [Andrew-node-1] timeout (30046ms) on waiting for synchronous responses from nodes=[Andrew-node-1, Andrew-node-2] responsesSoFar=[Andrew-node-1] request=id=2 task=resynch [OHazelcastDistributedDatabase]

[Андрей Логинов]

Issue has been raised https://github.com/orientechnologies/orientdb/issues/2514 .

[galina manashirova]

I have the same issue with 1.7.4.
Is there any work around before issue gets fixed?
Should I go back to the old version 1.6.4 where replication worked?