ozgune

20 ms·

181.
by ozgune 11y ago
Postgres in the Cloud
182.
by ozgune 11y ago
Data Analytics
183.
by ozgune 11y ago
New Technologies
184.
by ozgune 11y ago
Performance Tuning
185.
by ozgune 11y ago
Scaling Out for Big Data
186.
by ozgune 11y ago
Replication, Backup & Operations
187.
by ozgune 11y ago
Postgres in the Cloud
188.
by ozgune 11y ago
Data Analytics
189.
by ozgune 11y ago
New Technologies
190.
by ozgune 11y ago
So excited about upserts in Postgres! This was the most requested feature in PostgreSQL: https://postgresql.uservoice.com/forums/21853-general
191.
by ozgune 11y ago
(Ozgun from Citus Data) Sure, we'd be interested in running more numbers. If you have example data sets in mind, could you share them with us? For clarification, we picked this data set for several reasons. The data set was real, sizea
192.
by ozgune 11y ago
(Ozgun from Citus Data) Yes, that's a great question. The reason we didn't respond earlier was that we don't have a firm answer yet. Open sourcing CitusDB is a big decision for us, and we're still learning about (the bus
193.
Open source Windows? The unthinkable is already happening, says Microsoft(pcworld.com)
5 points by ozgune 12y ago|0 comments
194.
by ozgune 12y ago
(Ozgun from Citus Data) In the blog post, I used the + operator and sum() aggregate function interchangeably to be brief. Actually, those two operations are related, but have different representations in distributed relational algebra. I up
195.
by ozgune 12y ago
We'll update our FAQ with a detailed answer to this question. As a summary, the user specifies the shard and replication count as they are sharding their table. For example, if you have 4 nodes, you may pick 256 as the initial shard co
196.
by ozgune 12y ago
You don't need to worry much about data locality with pg_shard. It's designed to push the computation to the data, and as such, abstracts away most data locality concerns. The only exceptions are complex analytic workloads such as
197.
by ozgune 12y ago
(Ozgun from Citus Data) This benchmark confuses CitusDB with PostgreSQL + cstore_fdw extension. CitusDB scales out PostgreSQL to multiple machines, and cstore is a columnar store for PostgreSQL. The author has a clarification posted at the
198.
by ozgune 12y ago
Yes, huge congrats guys! I don't know if I love the product or the team more. :)
199.
by ozgune 12y ago
For clarification, it wasn't anyone from our team who downvoted any of these threads. All we can do on our part is to keep working on PostgreSQL to make it even better and faster, and to share our work with others in the community.
200.
by ozgune 12y ago
Yes, it is. Can started by trying these ideas on Postgres' row store, and he thought that we could get good benefits particularly when columns in the database table had fixed-widths. Then again, making changes for both fast column proj
201.
by ozgune 12y ago
(Ozgun from Citus Data here.) I updated the Readme, and added a second link to the original vectorization paper. We found that most benefits came from simply switching to batch-processing; and that's why we referenced the more recent &
202.
by ozgune 12y ago
(Ozgun from Citus Data here) We're actively working on improving the cstore_fdw extension and your feedback is valuable to us. We're keeping a list of open issues on GitHub, and prioritize them for upcoming releases based on your
203.
by ozgune 12y ago
Do these DAU/MAU numbers include WhatsApp users?
204.
Stephen Colbert: “We’re going to prove that I can sell more books than Amazon”(gigaom.com)
2 points by ozgune 12y ago|0 comments
205.
by ozgune 12y ago
I think part of the problem with referrals is that you want to know who among your employees' contacts are the better developers. If you pay for your employees' contact lists, how do you distinguish the stronger developers (or do
206.
by ozgune 13y ago
That sounds like a pretty good summary. From a workload perspective, row-stores are predominant in the OLTP (transactional insert/update) domain. They are also used in OLAP and data warehousing. Still, column stores have benefits when
207.
by ozgune 13y ago
Nope, you don't need to purchase CitusDB. We open sourced cstore_fdw with the Apache license. You can just build and load the extension, and start using it with your PostgreSQL 9.3+ database. If you later need to scale out your Postgre
208.
by ozgune 13y ago
(Ozgun from Citus Data here.) We're going to publish more comprehensive results in a few weeks. In our initial tests, we found the speed up to depend a lot on the underlying data and the type of queries. For example, we found that when
209.
by ozgune 13y ago
(Ozgun from Citus Data here.) Foreign tables provide a nice abstraction to separate the storage layer from the rest of the database. They are similar to MySQL storage engines in a sense. With them, users don't need to make any changes
210.
PostgreSQL old data structure: the HyperLogLog(github.com)
1 points by ozgune 13y ago|0 comments

More ›