IBM Netezza Performance Server Reviews

Enterprise reporting data warehouse using Business Objects, Microstrategy and data mining using SQL. Being a data repository for a single customer view. Also contained staging tables, some of which were designed like an ODS and contained all data from the source system and was updated on a nightly basis. The applicance contained over 12 TB of data uncompressed (less than 4 TB compressed).

Reports which used to take between 30 to 60 minutes or would time out on an Oracle database, which was previously used for the enterprise DWH, now run consistently in seconds or in less than five minutes.

High performance RDBMS appliance optimized for data warehousing and enterprise reporting. Very simple to manage huge volumes of data without having to worry about indexing and partitioning. Automated compression of tables without any custom scripting or manual intervention. Achieved almost 3x compression effortlessly which meant that 12 TB of data compressed into around 4 TB.

Could do better to support more concurrent update queries. We had to stagger our ETL loads to prevent queuing of jobs and random failures.

Also, it would have been good if the admin application showed more detail on the validity and usage of zone maps (this may have been implemented in later versions of the admin app).

The database runs stable unless there are hundreds of queries running in parallel.

Every query is a full table scan. If the table contains mostly integers, then performance is good. If the number of users is in the thousands, then it may be better to use cubes or other solutions to service reporting needs.

Before being acquired by IBM, Netezza corporation had exceptional support and used to respond very quickly (less than 30 minutes) in case of production issues. Round the clock support and monitoring were offered and support tickets were handed over very professionally between engineers working across time zones. After being acquired by IBM, support has not been as responsive, but there weren't as many issues as the box was stable.

Previously, Oracle was used as the data warehousing platform, and performance was low and not meeting the needs of the enterprise reporting and analytic user community. My customer switched to Netezza mainly for performance, and it was a big improvement.

As the box was very heavy, datacenter flooring required additional reinforcement. The box runs Linux and the initial setup is quite straightforward. ODBC drivers on the servers (ETL or reporting) which connect to the box may need to be upgraded.

Implemented this through a vendor team. As there is no need to spend time on partitioning and indexing, a lot of vendor time was saved. Table scripts for partitioned oracle tables run into hundreds or thousands of lines of code and we used to be charged accordingly. But a Netezza table script is much much simpler and we saved money there. Review of table scripts for performance and best practice was also easier as there is only a limited set of best practices to be implemented for high performance. So even vendor teams having low or medium level of expertise can deliver properly as long as they understand how MPP works - governance effort is definitely lesser with Netezza compared to Oracle or SQL Server.

ROI is high because analyst productivity improved drastically. As mentioned before, queries which used to run for several minutes now run in seconds or less than a few minutes or the duration of a typical pop song. So analysts can ask more questions of the data per hour compared to Oracle.

Also, the compression feature saved us a lot of money on per terabyte costs for the data.

From a cost per terabyte perspective, Netezza is definitely more expensive compared to Hive on Hadoop, but due to its simplicity and ANSI SQL Compliance and high performance which can be achieved with less tuning, it may be worth it.

My customer upgraded from Netezza 4.x to Twinfin 6.x.

Netezza is a great option for data warehousing, but give due attention to concurrency and find out how much would be the peak load the database may have to handle. Also, check whether performance is acceptable for APIs and web services. Performance may not scale for thousands of single row lookups, as the database is more suited for complex aggregated data warehousing queries.

Currently, we are using Netezza to the utmost. We first used it for our data warehouse. Then we moved on to doing analytics. Also, we are now doing some packages on it. Currently we are using it for multiple purposes, but, mostly it's used for reporting.

It is a back end for our SSIS, MicroStrategy,, Tableau. All of these are connecting to get the data. To do so we are also using our analytics which is built on the data.

I'm working for a retail client. We have a lot of reports, which are for store managers. The store managers used to get all the reports using the SSIS package, which was built upon our back end and they would have a lot of performance issues. When I say performance issues, in the morning when there are almost 2,000 users trying to connect to the same report, there were a lot of problems because it was a relational database. People were trying to get information out of it and it would break the flow, actually bottleneck the flow. That's why they were getting the impact.

What we did is, we did a PoC and found this client is well built. Because of that, we were able to create a separate layer for reporting. There was a simple requirement for the DBA, where we don't need to build an index or anything like that. Those reports were improved from a couple of hours to minutes. That was the biggest benefit we've had from this appliance.

For me, mainly, it reduces my costs. It's not only the appliance cost. There are also support costs and a maintenance costs. It does reduce the costs very drastically.

Administration of this product is too tough. It's very complex because of the tools which it's missing. We would require better tools for doing administration. For example, if I need to get a permission of a user on a particular object, it is so complex. It's not straightforward. It would not give me a display of that particular element. Basically auditing is tough to do on it. If I don't have a graphical tool, the auditor will ask me thousands of questions.

This is my seventh year.

Stability, no. In production, no. But when we are migrating or we are moving out from one platform to another... but I guess that doesn't count.

It is scalable to multiples of itself. If currently it is 1X, I can go to 3X, 4X, 5X. I cannot go 1.2, 1.5, 1.8. I have to go 1X, 2X, 3X, a whole interval. That is the only catch, otherwise yes, it is scalable.

We were using a different product. We switched it because of the cost of maintenance. We wanted to reduce the cost of maintenance and it helped to drastically reduce it by some 40 to 50%.

I'm not sure what to compare it to. I am a DBA. For me to set it up, it would be a fast setup. If I'm not a DBA, then it would totally take a little bit of time. For someone who is new to Netezza, it's not complex but it requires prerequisite knowledge. That's it.

I would rate the setup as medium complexity, given that we have somebody who knows a little bit of database administration. I would not say the full administration. I would rate setup around two and a half out of five.

We did. We were actually looking at Teradata and Exadata and we chose Netezza.

The first thing is, do a full PoC where you actually plug in dummy data. Start with millions of records and try out the options which you have already in your environment and see if you really get a benefit in terms of performance. Because this product is mainly to improve the performance.

There are two things. First, it improves the performance and then it's easy to maintain. When I say easy, you would not need a DBA at all. Anybody who has really good knowledge in SQL should be able to maintain this product. That cost, of a DBA, you will completely be eliminating.

The first part of this product is basically performance. If you are moving on from an ODBC database to this, then you would need to do a round of testing on data because you are actually moving from a different technology, that is, to an appliance technology. There are multiple things which can stop you, things you might not be able to do; operations, like booking of a ticket or reloading a feed. We tried to do those things here, in Oracle or MySQL, but when it comes to reporting, this is the best solution we have so far. By that I mean, cheapest and the best.

Then, try to figure out if you really require a code change, because there is a minimal time for a code change; but if it's a request then you should catch it when you're doing a PoC.

Take out any complex situation, give it to the vendor, IBM. Tell them to do a small PoC on the dummy data and let them come up with the performance, then you can compare with your feeds.

Overall, it has good connectivity with almost all the niche technologies we have on the market right now. It can scale up to any technology, any content. That is another benefit of having this.

IBM is a good company which always tries to be competitive with other technologies on the market. If we are stuck, we are trying to purchase something for our content and we are stuck somewhere, something is missing, we approach them and these guys will give us the solution in terms of a patch or an API or something like that, which will help us connect to the newest technologies.

• Its ability to process and query large amounts of structured data.
• Low administrative support in terms of query optimization and indexing support. Indexing and data partitioning is built into the firmware.
• Data compression. It was relatively impressive. I think at some point we were getting 4:1 compression if not more.

I can't because we're using it in a fashion such that a traditional RDMS could have been used in place of it. Our data was relatively small so we didn't see a huge benefit in transitioning new subject areas into production.

Community support. There was none, or very little, especially when using add-on software (e.g. building functions, MapReduce, R, Lua, etc.).

Driver support for windows based applications.

Disaster recovery support. Because it was an appliance, and if you wanted to support disaster recovery, you needed to buy two.

Data center moves. The TwinFin was cool to look at, but moving was a huge orchestration. I would definitely go cloud if I was to choose something with this type of processing power.

Four years.

Yes, with drivers, either OLE DB or ODBC connectivity. Poor backwards compatibility and performance issues with subscribing applications, including MicroStrategy.

No, we didn't have enough data to really press it. The TF-6 at the time could support towards 64TB. We were using 10TB. When running queries natively, it processed most large data queries quickly.

Netezza promoted MapReduce functionality as add-on software, but I thought the support for it was very poor. I wasn't aware of any community that was able to get that up and running.

• Speed
• storage
• RAM

all of which contribute to large capacity.

Faster data processing compared to commodity servers.

In-DB processing with SAS Analytics, since this is supposed to be an analytics server so the expectation is there.

Three-plus years.

No.

No.

Eight out of 10.

Unix, commodity server. Switched to check this technology.

SAS server (Unix) and data storage only.

Expensive to maintain compared to other solutions.

Teradata and Greenplum.

Make sure to check the capability regarding in-DB processing with the application you will be using with this appliance. I am using SAS and it's a dis-appointment due to limited in-DB processing, including data connection pool, that was never resolved by both SAS and IBM Netezza.

It is extremely easy to use. The platform is very efficient and robust as well.

Generally, the machine utilized for the PoC is rarely returned once the overwhelming results are seen! Customers decide to keep the project, almost from scratch! The initial KPIs are delivered with low investment. There is quick implementation and considerable outcomes which make the platform a valuable asset to the organization.

Marketing, I dare say, should be improved. The ads should never mention “big data” since the platform was not initially designed for it.

I worked years and years ago with Teradata, and this is the platform to which it should be compared. It is important to emphasize that Netezza and Teradata have different markets. Netezza performs very well “in its peculiar and valuable small world.”

I have used this solution for years.

We did not have any stability issues.

It is not designed to process a huge volume of data.

Technical support is good.

I worked for different companies and my experience is really wide. Teradata is the paramount solution! Oracle’s Exadata invests a lot in both technology and marketing, sometimes losing the focus on “real” analytics. They used to address the platform as both OLAP and OLTP (“one size fits all”), which is not correct.

The rule of thumb in the data warehousing/business intelligence space starts with a simple PoC. That being said, there must be a few questions to be answered by the platform. The data model must be simple and flexible enough to address some “extra questions” within a set timetable, in case the customer wants to extend the PoC prior to making the purchase.

Comparisons with other solutions, including pricing and licensing, is as important as a PoC.

I compared Netezza to Teradata and Exadata.

IBM PureSystems or NPS has a patented NDAC or Netezza data accelerator card inside each of the SPUs (Snippet Processing Arrays). The SPUs act as individual processors that share part of a query. This helps in high-speed parallel processing of big data volumes. Querying a large data volume takes just a few seconds to minutes, if it is done right.

Analyzing years of data requires high processing power and storage. IBM PDA has exactly that. Years of processed data (tables) can be queried and retrieved based on management requirements. This can be done in minutes for analysis. This is extremely important in identifying trends for decision making in higher management, to serve customers better in today’s business environment.

It is a highly complicated architecture and only IBM engineers/support, or someone who worked on the hardware side of the system, can understand the system architecture completely.

This means that:

• Replacements can only be done by an IBM engineer
• Components are not generic and are not on the market
• Many of the systems are IBM patented
• Service support costs are high

Storage, although high, is limited depending on the rack configuration. For example, an N3001-80, which is an eight-rack Mako Server and also the top end model with the highest storage, can store a maximum of 384TB. Should your data exceed this limit, the storage cannot be extended.

It cannot be used for unprocessed data. The data has to be in a table format.

We have been using the solution for five years.

We rarely had stability issues. You may expect an average of one outage every two months for a few minutes, if it is maintained properly. This is mostly due to “pollreplytimeout” errors of the SPU.

All the rack components are dual for redundancy. Almost all components have failsafe/backup, including the host that is configured in cluster for high availability. With this, the system is up and running all the time, or you can get it up at the earliest possible time in case of component failures/system down situations.

We had scalability issues. Storage, although high, is limited depending on the rack configuration. For example, an N3001-80, which is an eight-rack Mako Server and also the top end model with the highest storage capacity, can store a maximum of 384TB. Should your data exceed this limit, the storage cannot be extended. You will have to purchase a new appliance.

I would rate the technical support at 7/10.

We did not use a previous solution.

IBM sets up the NPS and does the initial configurations. As an admin, you will have a completely configured system ready to work on.

This is a multi-million dollar product. The software and hardware are both IBM patented.

Netezza SQL ('nzsql') runs only on PDA systems, and cannot be installed on other enterprise-class servers. Power consumption is high and these systems must have mandatory IBM support, due to the patented and exclusive IBM hardware. However, the benefits outweigh the costs if you have large data volumes that require fast analysis on a day-to-day basis.

We didn't really examine any alternatives. I started working directly on IBM PDA (Netezza) and I haven't had any chance to work on the competitors' products. I was not a decision maker in the selection of this product. This was a higher management decision.

A close competitor of IBM PDA is Apache Hadoop. As of now, no other product is as fast or as stable.

You can definitely consider this appliance if you have:

• A large volume of processed data (tables) that are created on a daily basis
• Data that requires daily analysis, critical for decision making, and a budget to complement it.

This is one of the most stable and fastest data warehouse appliances available in the market today.

Distribution (no index or partition), built-in analytics is a major advantange.

It has improved our management information systems and reporting capabilities.

It should also consider cloud based solutions.

We have been using the solution for more than ten years from the 8000 series,

We did not encounter any issues with stability.

I can't extend the storage, only up to 6x compress of data. You need to plan this when selecting the right product to buy.

Netezza support is good.

We used Oracle previously. The performance required less maintenance when comparing index and data maintenance issues.

The setup took only a few hours with all built-in apps. It is very quick and easy to use.

One license and one support, also cheaper compared to other products.

We evaluated MS SQL server.

It is easy to use. Make sure you select the right ETL and reporting tool. Also select the right tool for the organization to hold it in the long run.

It has a compression engine and FPGA on but you should still analyze your volume of data and decide on the right model and size.

We use Aginity to access Netezza database. We really like the way we can dominate the physical distribution of the data hence know how to improve the performance of the query. Netezza on its own is very fast which makes our life easier to run huge queries for analytics.

This has really helped us to improve the performance of our Data Marts and warehouses. We can run our reports very quickly. It has also improved the turn around time of business requests. 

In Aginity there should be a way to format the SQL queries. I think we can't format the query the way we can do it in Oracle editor (beautifier). Say, for example, if we are trying to get the DDL of an existing view, we lost the formatting. It's a minor issue, but important from usability point of view. Other than this, I think we are good so far with Netezza as a whole.

We have been using this product from last 5 years. 

We did not experience any issues.

We haven't had any issues.

We've not experienced any issues.

If volume is the issue, use Netezza. Nothing is better than this product.