-- Channels
insert into sym_channel
(channel_id, processing_order, max_batch_size, enabled, description)
values('announcement', 1, 100000, 1, 'announcement info');
-- Node Groups
insert into sym_node_group (node_group_id) values ('ORA');
insert into sym_node_group (node_group_id) values ('POST');
-- Node Group Links
-- oracle sends changes to postgres when postgres pulls from oracle
insert into sym_node_group_link (source_node_group_id, target_node_group_id, data_event_action) values ('ORA', 'POST', 'W');
-- postgres sends changes to oracle when postgres pushes to oracle
insert into sym_node_group_link (source_node_group_id, target_node_group_id, data_event_action) values ('POST', 'ORA', 'P');
-- Triggers
-- Triggers for tables on channel_id
insert into sym_trigger
(trigger_id,source_schema_name,source_table_name,channel_id,last_update_time,create_time)
values('announcement','OPENVOICEED1','ANNOUNCEMENT','announcement',current_timestamp,current_timestamp);
-- Routers
-- Default router sends all data from oracle to postgres
insert into sym_router
(router_id,source_node_group_id,target_node_group_id,router_type,create_time,last_update_time)
values('oracle_2_postgres','ORA', 'POST', 'default',current_timestamp, current_timestamp);
-- Default router sends all data from postgres to oracle
insert into sym_router
(router_id,source_node_group_id,target_node_group_id,router_type,create_time,last_update_time)
values('postgres_2_oracle','POST', 'ORA', 'default',current_timestamp, current_timestamp);
-- Trigger Routers
insert into sym_trigger_router
(trigger_id,router_id,initial_load_order,last_update_time,create_time)
values('announcement','oracle_2_postgres', 100, current_timestamp, current_timestamp);
insert into sym_trigger_router
(trigger_id,router_id,initial_load_order,last_update_time,create_time)
values('announcement','postgres_2_oracle', 100, current_timestamp, current_timestamp);
Hi, I had google and double checking engine file and registration. i still facing the following problem. Basically i duplicate the H00002 and rename and change all the setting become H00004. Both engine files sit in a same server while the db is pointing to RDS. Before I register H00004, i had double check sym_node, sym_node_host and sym_node_security and there is no H00004 record. After i register H00004 and start my client symmetric with H00004 engine file ONLY. It return me the following error.
~~~Error log~~~~~
2019-07-18 23:19:31,988 ERROR [corp-H00004] [AbstractSymmetricEngine] [symmetric-engine-startup-1] An error occurred while starting SymmetricDS StackTraceKey.init [SymmetricException:2675662801]
org.jumpmind.symmetric.SymmetricException: The configured state does not match recorded database state. The recorded external id is 'H00002' while the configured external id is 'H00004'. The recorded node group id is 'corp' while the configured node group id is 'corp'
at org.jumpmind.symmetric.AbstractSymmetricEngine.start(AbstractSymmetricEngine.java:658)
at org.jumpmind.symmetric.AbstractSymmetricEngine.start(AbstractSymmetricEngine.java:634)
at org.jumpmind.symmetric.ClientSymmetricEngine.start(ClientSymmetricEngine.java:271)
at org.jumpmind.symmetric.web.SymmetricEngineHolder$EngineStarter.run(SymmetricEngineHolder.java:598)
at java.util.concurrent.ThreadPoolExecutor.runWorker(Unknown Source)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(Unknown Source)
at java.lang.Thread.run(Unknown Source)
I even compare the engine file using beyong compare.
Need your advise where i need to check?
I like to sync tables within the same database on event trigger, from incoming SymmetricDs batch or even a column change from another application. I read this is not possible, because the source and targe nodes are the same. Is there way/setting to workaround this? Thanks in advance.
I have an issue on a table reload requests.
I use this sql: insert into sym_table_reload_request (TARGET_NODE_ID, SOURCE_NODE_ID, TRIGGER_ID, ROUTER_ID, CREATE_TIME, LAST_UPDATE_TIME, CREATE_TABLE, DELETE_FIRST) VALUES ('erikvis_webdb_local', 'erikvis_webdb', 'artikel_web',' webdb2webdb_local', current_timestamp, current_timestamp, 1, 1);
When I use 'ALL', 'ALL' for "trigger_id" and "Router_ID", I have the same issue.
Yes I have vice versa.
insert into sym_node_group_link (SOURCE_NODE_GROUP_ID, TARGET_NODE_GROUP_ID, DATA_EVENT_ACTION, SYNC_CONFIG_ENABLED, IS_REVERSIBLE, CREATE_TIME, LAST_UPDATE_BY, LAST_UPDATE_TIME) VALUES ('erikvis_webdb', 'erikvis_webdb_local', 'P', 1, 0, current_timestamp, 'CWS', current_timestamp);
insert into sym_node_group_link (SOURCE_NODE_GROUP_ID, TARGET_NODE_GROUP_ID, DATA_EVENT_ACTION, SYNC_CONFIG_ENABLED, IS_REVERSIBLE, CREATE_TIME, LAST_UPDATE_BY, LAST_UPDATE_TIME) VALUES ('erikvis_webdb_local', 'erikvis_webdb', 'P', 1, 0, current_timestamp, 'CWS', current_timestamp);
now that i'm synching data to both a mysql and an mssql host i cant find a way to get initial.load.delete.first=true to work, the previous workaround of setting initial.load.delete.first.sql=delete from \"%s\" breaks the mysql side where it seems that escaping is already being done?
Hi, I have a production environment with up 3 layers of communication store -> region (aws) -> HQ (aws). Our DB is Ms SQL and install at RDS.
I also running symmetric clustering on region layer with 2 region server (each region server have 5 engine file to handle different region request).
Both region server also have network load balance based on AWS setting. At first, the clustering is running fine until last 2 week ago, the data somehow duplicated.
The actual table PK is just an ID identity. And we are using sym_trigger sync_key_names to determine the actual PK for the table. The same table we also have
insert and update action. When the incident happen, a same line of record it had duplicate 10 to 20 times. May I know how the symmetric clustering work? or any
advise for us to take note? We are using symmetric version 3.9.4.
We also have another RDS performance issue when we start all the symmetric. RDS CPU, High IOPS usage when we start up the symmetric.
Currently, we had upgrade the RDS spec to a very high performance server but the CPU usage still remain very high. To minimize the problem,
we had econfigure our symmetric engine
http.concurrent.workers.max=12 (for the region engine),14 (for HQ engine)
purge.retention.minutes=4320
FYI, we will end up have 8 region which going to manage 150 stores data sync up.
However, lesser threads it will cause the data flow in from store to region become very slow.
For this case, is there anything we can further fine-tune?
