Pivotal Greenplum Review

The MPP element is crucial, so far as it allows us to query millions of rows at a time, at speed.


What is most valuable?

The MPP element is crucial, so far as it allows us to query millions of rows at a time, at speed.

How has it helped my organization?

The previous data warehouse was built in Oracle. One of the things which has improved in GreenPlum is that we can query millions of rows at speed, without creating lags. We’ve also built far more views; slowly changing dimensions can instantaneously update without creating the issue of having to rebuild tables to reflect new hierarchies, for example.

What needs improvement?

We found some issues with larger tables that have daily data appended, where after a while this seems to create lag in the query speed. This might just have to do with local knowledge rather than the product itself.

We have a table which is currently contains 27.6m rows and has a daily delta added to it of roughly 16.5k rows per day. While this isn’t particularly large, we have noticed the table begins to perform poorly when queried, in spite of having set up a VACUUM process to be performed weekly. It may be that the VACUUM process needs to be performed more frequently (like daily), but we’ve not yet found the optimal way of maintaining this particular table.

It’s worth saying that this is one table out of over 400 perfectly well performant tables and views in the same database. Hope that helps,

For how long have I used the solution?

I have used for approximately 30 months.

What was my experience with deployment of the solution?

I have not encountered any deployment, stability or scalability issues.

How is customer service and technical support?

I have not raised any service issues/tech queries, so I can’t really say.

Which solutions did we use previously?

We used Oracle previously. We based our choice on expertise in our US operation, where we have a GreenPlum expert who provided some amazing use case examples to help us in our selection process.

What about the implementation team?

Implementation was done in-house.

What was our ROI?

Not within my area I’m afraid, but I understand that this was a very good fit from an ROI point of view

What other advice do I have?

Investigate whether this solution works for you. It is worth creating a rating matrix to compare other similar products, and it is very useful to look deeply at whether the third-generation MPP software might be a good fit.

Disclosure: I am a real user, and this review is based on my own experience and opinions.
1 Comment
author avatarHimanshu Paliwal
Real User

Regarding the performance of few large tables, just a suggestion you can also try implementing the partitioning. By doing partitioning you can leverage the "swap partiton" while doing an insert and select the data for reporting based on your partitioning key.
Hope this helps

Guest
Sign Up with Email