Activity for Neil Clayton
-
Neil Clayton posted a comment on discussion Help
I'm setting up a scenario where each customer has some shared data (currently residing in the 'public' schema) and their own data in their own named schema (e.g: site_wft_1). This is on postgres. The tables are generated via Django, and contain FK links from the per-'User' models to the 'Shared' models. e.g: site_wft_1.System4000_User_job references public.System4000_Shared_customer When I try to do a load, the creation of the 'users' tables fails because the schema naming is incorrect. In the logs...
-
OK. Here is some DB SQL: https://www.dropbox.com/s/1o84qpi3bpinf2c/symmetricDSArchive.zip?dl=0 This includes various SQL files you might find useful. I dumped the entire DB after setting it up (including sample data and triggers/routers). See "Entire Master.sql". I can replicate the problem by executing the DeleteTheLogin script on master, and seeing the fault on a slave (well, seeing it not work because the router can't make the link to the login any more). I've included properties files for both...
-
Will do. It'll take me a while to par it down to something simple. I'll see if I can get a simple self contained project available to demonstrate the prob.
-
Will do. It'll take me a while to par it down to something simple. I'll see if I can get a simple self contained project available to demonstrate the prob.
-
I've been testing SDS as a solution for the master-corp/separate stores situation, where each store has a subset of data. I'll use the term "master" to refer to the master-db. The 'corp'. I'll use the term "client" to mean any node with filtered data. I'm running into two problems: 1) If I delete both a m2m row and dependent row, the dependent row isn't deleted on the client. 2) If I update the owner of the row, the old dependent row isn't deleted on the old client (the previous owner) Triggers use...
-
I've been testing SDS as a solution for the master-corp/separate stores situation, where each store has a subset of data. I'll use the term "master" to refer to the master-db. The 'corp'. I'll use the term "client" to mean any node with filtered data. I'm running into two problems: 1) If I delete both a m2m row and dependent row, the dependent row isn't deleted on the client. 2) If I update the owner of the row, the old dependent row isn't deleted on the old client (the previous owner) Triggers use...
-
I've been testing SDS as a solution for the master-corp/separate stores situation, where each store has a subset of data. I'll use the term "master" to refer to the master-db. The 'corp'. I'll use the term "client" to mean any node with filtered data. I'm running into two problems: 1) If I delete both a m2m row and dependent row, the dependent row isn't deleted on the client. 2) If I update the owner of the row, the old dependent row isn't deleted on the old client (the previous owner) Triggers use...
-
I've been testing SDS as a solution for the master-corp/separate stores situation, where each store has a subset of data. I'll use the term "master" to refer to the master-db. The 'corp'. I'll use the term "client" to mean any node with filtered data. I'm running into two problems: 1) If I delete both a m2m row and dependent row, the dependent row isn't deleted on the client. 2) If I update the owner of the row, it isn't deleted on the client Triggers use an EXTERNAL_SELECT to fill in EXTERNAL_DATA....
-
I've been testing SDS as a solution for the master-corp/separate stores situation, where each store has a subset of data. I'll use the term "master" to refer to the master-db. The 'corp'. I'll use the term "client" to mean any node with filtered data. I'm running into two problems: 1) If I delete a row, it isn't deleted on the client. 2) If I update the owner of the row, it isn't deleted on the client Triggers use an EXTERNAL_SELECT to fill in EXTERNAL_DATA. They do this using joins. In this simple...
-
I’ve got SDS (3.97) working for filtered updates to slave nodes. My scenario is identical to the example in SDS… the master with various stores. In my case, I’m using external_select on each trigger, to insert the "owner name" for each modified row. The router in use uses BSH to check that the external ID begins with that name. I’m wondering how I might do an initial load, that sends only the filtered data, to each slave node. Currently nothing comes through, I think because the reload doesn’t add...
-
Answering my own question. It is because my registration.url and sync.url were subtly different. Thus, isRegistraitonServer() returned false. I made registration.url= (i.e: nothing... blank). This change was made only on the server side, of course. And ta daaa. It now works.
-
So, I thought I had this all going. But alas, I can't seem to get an initial load extract to even start on the source. Here's the setup: Source is masterfoo. Single destination is PORIRUA-1 There is data in the Organization and Product tables (referenced below). Nodes INSERT INTO `sym_node` (`node_id`, `node_group_id`, `external_id`, `heartbeat_time`, `timezone_offset`, `sync_enabled`, `sync_url`, `schema_version`, `symmetric_version`, `config_version`, `database_type`, `database_version`, `batch_to_send_count`,...
-
Thankx Max - I ended up using unique external-ID's with a shared prefix, then using a bsh router to say "route to that if it begins with that prefix". for (org.jumpmind.symmetric.model.Node node : nodes) { if (ORGANIZATION_ID != null && node.getExternalId().startsWith(ORGANIZATION_ID.trim())) { targetNodes.add(node.getNodeId()); } }
-
Hi! I seem to have SDS working (auto.registration for now). I’ve set it up using 3 separate docker containers/dbs, and it replicates from master -> slave and back again (using two routers). Initially I setup using 3 separate external IDs. e.g: 1, 2 and 7. I then wanted to setup 2 hosts with the same external ID. Lets use ‘2’ in this example. I can register the first instance, but upon starting the 2nd instance, it tried to register with what looks to be the same node_id (2). Client1 properties: engine.name=WFTClient...
-
Hi! I seem to have SDS working (auto.registration for now). I’ve set it up using 3 separate docker containers/dbs, and it replicates from master -> slave and back again (using two routers). Initially I setup using 3 separate external IDs. e.g: 1, 2 and 7. I then wanted to setup 2 hosts with the same external ID. Lets use ‘2’ in this example. I can register the first instance, but upon starting the 2nd instance, it tried to register with what looks to be the same node_id (2). Client1 properties: engine.name=WFTClient...
1
