I wanted to understand how the two projects are related or will evolve - Seems like this project is not under as active development as the Postgres-XC project which seems to have some of the folks who developed the GridSQL software on it. Postgres-XC seems to target a wider problem but from an initial look at the documentation for Postgres-XC it seems to be support many of the features that GridSQL provides (and more). Even the documentation style seems similar perhaps due to the overlap of the team members working on both projects.
Would it be possible for someone using GridSQL to move to Postgres-XC easily? What state is the Postgres-XC project in with regard to SQL support etc. Is Postgres-XC a planned direction or will this project be actively developed for some time into the future?
Thanks
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hello,
I can understand the confusion. You are absolutely correct, that the XC project is very active, but it has a very different goal than GridSQL. Also, the GridSQL project is currently in a transition phase. New development is now being done at https://launchpad.net/stado with the project being called Stado. One can very easily move from GridSQL to Stado since it is the same code base with simply a new name.
The XC team is working to have a horizontally scalable version of PostgreSQL that scales for transactional workloads while Stado is aimed at business intelligence work loads. The big difference is the use of parallelization across the nodes. Currently XC does not parallelize the queries across the nodes, but Stado does. In the future, XC may develop the same features as Stado, and I would suspect that it would be an easy move if the data uses the same partition key in both solutions.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Thanks for the quick reply - Do you know if this project is going to be maintained or whether Stado is what anyone (on GridSQL who is looking to get any updates/bug fixes etc) should migrate to? Also, do you have information on how many developers are working on the Stado team?
Is this project officially inactive or transferred to Stado? - Don't see anything on this site to indicate it is closed or no longer under development.
Thanks.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Some of those exact details are still being worked out, but since EDB has stopped all support for GridSQL, myself and others have moved over to Stado where we will be actively maintaining it and adding new features. I don't know of anyone who plans on supporting the GridSQL code base so my personal recommendation is look to Stado for bug fixes.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Postgres-XC currently focuses on OLTP, not DW/BI/OLAP. Postgres-XC is currently limited in its ability to handle complex queries efficiently in a parallel manner, whereas GridSQL does this well.
Conversely, Postgres-XC will perform much better for OLTP type of workloads.
Niether supports all PostgreSQL functionality 100%. I would say that GridSQL does a better job at more generic SQL right now, but that will likely change, and Postgres-XC does a better job with PostgreSQL-specific functionality and will continue to improve.
I also want to add that, according to EnterpriseDB (I am a former employee), they will no longer officially support it:
I figured I could further add to the confusion and mention that users looking to GridSQL or Stado for BI/DW should instead take a look at Postgres-XL. There is less latency and better SQL coverage, it is a modified version of PostgreSQL directly, instead of a separate layer like GridSQL or Stado.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi,
I wanted to understand how the two projects are related or will evolve - Seems like this project is not under as active development as the Postgres-XC project which seems to have some of the folks who developed the GridSQL software on it. Postgres-XC seems to target a wider problem but from an initial look at the documentation for Postgres-XC it seems to be support many of the features that GridSQL provides (and more). Even the documentation style seems similar perhaps due to the overlap of the team members working on both projects.
Would it be possible for someone using GridSQL to move to Postgres-XC easily? What state is the Postgres-XC project in with regard to SQL support etc. Is Postgres-XC a planned direction or will this project be actively developed for some time into the future?
Thanks
Hello,
I can understand the confusion. You are absolutely correct, that the XC project is very active, but it has a very different goal than GridSQL. Also, the GridSQL project is currently in a transition phase. New development is now being done at https://launchpad.net/stado with the project being called Stado. One can very easily move from GridSQL to Stado since it is the same code base with simply a new name.
The XC team is working to have a horizontally scalable version of PostgreSQL that scales for transactional workloads while Stado is aimed at business intelligence work loads. The big difference is the use of parallelization across the nodes. Currently XC does not parallelize the queries across the nodes, but Stado does. In the future, XC may develop the same features as Stado, and I would suspect that it would be an easy move if the data uses the same partition key in both solutions.
Hi,
Thanks for the quick reply - Do you know if this project is going to be maintained or whether Stado is what anyone (on GridSQL who is looking to get any updates/bug fixes etc) should migrate to? Also, do you have information on how many developers are working on the Stado team?
Is this project officially inactive or transferred to Stado? - Don't see anything on this site to indicate it is closed or no longer under development.
Thanks.
Some of those exact details are still being worked out, but since EDB has stopped all support for GridSQL, myself and others have moved over to Stado where we will be actively maintaining it and adding new features. I don't know of anyone who plans on supporting the GridSQL code base so my personal recommendation is look to Stado for bug fixes.
Hello,
Postgres-XC currently focuses on OLTP, not DW/BI/OLAP. Postgres-XC is currently limited in its ability to handle complex queries efficiently in a parallel manner, whereas GridSQL does this well.
Conversely, Postgres-XC will perform much better for OLTP type of workloads.
Niether supports all PostgreSQL functionality 100%. I would say that GridSQL does a better job at more generic SQL right now, but that will likely change, and Postgres-XC does a better job with PostgreSQL-specific functionality and will continue to improve.
I also want to add that, according to EnterpriseDB (I am a former employee), they will no longer officially support it:
http://forums.enterprisedb.com/posts/list/2618.page
However, there is a new fork of GridSQL called Stado that was recently announced. It includes a few bug fixes compared to GridSQL.
http://www.stado.us
With EnterpriseDB no longer supporting GridSQL, Stado could become the de facto successor.
Regards,
Mason Sharp
I see Jim beat me to a reply. :-)
Jim, Mason,
Thanks a lot for your immediate response
Regards
I figured I could further add to the confusion and mention that users looking to GridSQL or Stado for BI/DW should instead take a look at Postgres-XL. There is less latency and better SQL coverage, it is a modified version of PostgreSQL directly, instead of a separate layer like GridSQL or Stado.