artifact_id;status_id;status_name;priority;submitter_id;submitter_name;assigned_to_id;assigned_to_name;open_date;close_date;last_modified_date;summary;details;"Category";"Group" 1000247;1;"Open";3;100103;"Carlos Rocha";100;"Nobody";"2005-02-20 00:37";"";"2012-09-24 02:14";"Seems good stuff but...";"What about us, poor M$ Windows users ?";"None";"None" 1000384;1;"Open";3;100;"Nobody";100;"Nobody";"2005-09-15 04:44";"";"2012-09-24 02:14";"8.0.3 compatible";"When patching 8.0.3 source with the pgcluster patch for 8.0.1 I get some errors. I read in the mailing list that the patch should work with 8.0.3 but I don't want to compile a broken postgres in cause of this errors while patching. 1. If it should work with 8.0.3 => Why does this patch errors appear? 2. If it sould not work with 8.0.3 => Please provide a patch for 8.0.3 - pgcluster is the most powerfull cluster mechanism I have seen until right now, using it only with 8.0.2 or 8.0.1 is a big problem.";"None";"Next Release (example)" 1000393;1;"Open";3;100;"Nobody";100;"Nobody";"2005-09-25 13:11";"";"2012-09-24 02:14";"[Pgcluster-general] Asynchronous & Disconnected Multi-Muster";" Dear Mr Atsushi Mitani, hello Can you please clarify: is it difficult to make Pgcluster work in ""asynchronous & disconnected"" mode? This request is similar to http://pgfoundry.org/forum/forum.php?thread_id=265&forum_id=259 Probably, your team had some progress on this feature? Probably, we can help your team? For example, please consider this scenario: 1) say, we have 4 DB Servers (at location #1,2,3,4), 0 Load Balance Servers, 2 Replication Servers (at location #5,6) 2) then all 2 Replication Servers go down; or connection to a location #1 goes down, etc; this makes some or all 4 DB Servers DISCONNECTED 3) now what we want is to continue for all 4 DB Servers (at different locations) to operate in ASYNCHRONOUS (meaning independent of other DB Servers) read-write mode, so they all can accept inserts/updates locally (assuming that they are able to store all accumulated updates/changes in their local logs for later replication) 4) finally, when any 1 Replication Server comes back online, all previously accumulated changes on DB Servers 1,2,3,4 must propagate changes to each other and become in-sync. (serials consistency is not an issue, as our application uses UUID for everything to be used as a key) 5) Probably, there is a compromise solution exists, say put 1 DB Server and 1 Replication Server on the same computer at each location? Thank you very much for your time, Regards, Mr Andrei Pozolotin ";"None";"None" 1000669;1;"Open";3;100;"Nobody";100;"Nobody";"2006-06-19 07:25";"";"2012-09-24 02:14";"pgcluster for 8.1.4";"Hi, Do you intend to release a version of pgcluster for postgresql 8.1.4 (or which one is latest) soon? It's pretty unusual for such a popular project to stale for so many time. Or what else should I use instead of pgcluster (except slony)? Thanks!";"None";"None" 1000683;1;"Open";3;100;"Nobody";100;"Nobody";"2006-07-11 08:26";"";"2012-09-24 02:14";"Multiple interface support with automatic life chek";"For HA systems the support of multiple interfaces is essential. Also automtatic life check support and interface error detection is essential in that case. PgCluster should not use hostnames but IP addresses in configuration files.";"None";"None" 1010391;1;"Open";3;108881;"manish singh";100541;"Atsushi Mitani";"2008-05-04 10:47";"";"2012-09-24 02:14";"runtime recovery";"can you include a feature for runtime recovery solution for pgcluster v.1.9 mean is that when a DB Server goes down, and others are still listing and in read-write mode, what the need now is just start the gone DB server and get the psql prompt, just type the recovery command and the new database should getup and start providing requests to load-balancer queries.. ";"Interface Improvements (example)";"Next Release (example)" 1010803;1;"Open";3;137972;"Kamil Czajko";100;"Nobody";"2010-05-04 06:18";"";"2012-09-24 02:14";"Support for 8.4 and column based security";"Hello PGCluster developers Firstly your product is fantastic and we have been using it in a production environment for a few months now. However our managers are now requiring us to use per column security. I was wondering when it was likely that there would be a pg cluster available with postgres 8.4 support? Secondly we have had two ""crashes"" of the system after they run for several months where the replicator locks up and needs to be restarted otherwise the databases can no longer be connected to (hang indefinitely, never time out). We are not using the load balancer as it does not support SSL mode and the systems we integrate with require it. Unfortunately I do not have any debug information from these times as we were just trying get everything up and running again as quickly as possible. Any way please keep up the good work your project is highly appreciated! :) Kamil";"None";"Next Release (example)" 1010969;1;"Open";3;205155;"johnson hsu";100541;"Atsushi Mitani";"2010-12-29 13:50";"";"2012-09-24 02:14";"a way to increasing pgcluster speed";"Hi! I just install pgcluster in my company for weeks, It's perfect. Because of network traffic problem, pgcluster is hard to be a big cluster system. If there're ten cluster db in the system, when user request one time, the replicate must run nine times at lease. It will cause network jam. I've an idea, using brocasting method. When loadbalance server received a request, It make a brocasting to all clusters DB with data in it, and every culster db use a standard TCP to tell loadbalance server it had received or not. with this method, the best situation is just only one time data transfer, the effect of pgcluster will be the same to stand alone DB does while updating data, and it's will be suitable to be a cloud DB. Best Regards Johnson";"None";"None"