51–60 of 475 posts
When I was hiring data scientists for a previous job, my favorite tricky question was "what stack/architecture would you build" with the somewhat detailed requirements of "6 TiB of data" in sight. I was careful not to require overly complicated sums, I simply said it's MAX 6TiB I patiently listened to all the big query hadoop habla-blabla, even asked questions about the financials (hardware/software/license BOM) and…
This is not fully correct. Originally big data was defined by 3 dimensions: - Volume (mostly what the author talks about) [solved] - Velocity, how fast data is processed etc [solved, but expensive] - Variety [not solved] Big Data today is not: I don't have enough storage or compute. It is: I don't have enough cognitive capacity to integrate and make sense of it.
What do you mean by 'variety'?
One of the reasons big data systems took off was because enterprises had exports out of third party systems that they didn't want to model since they didn't own it. As well as a bunch of unstructured data e.g. floor plans, images, logs, telemetry etc.
When I was hiring data scientists for a previous job, my favorite tricky question was "what stack/architecture would you build" with the somewhat detailed requirements of "6 TiB of data" in sight. I was careful not to require overly complicated sums, I simply said it's MAX 6TiB I patiently listened to all the big query hadoop habla-blabla, even asked questions about the financials (hardware/software/license BOM) and…
You have 6 TiB of ram?
Everyone was talking about data being the new "oil".
Enterprise could basically deploy petabyte scale warehouse run HBase or Hive on top of it and build makeshift data-warehouses. It was also when the cloud was emerging, people started creating EMR clusters and deploy workloads there.
I think it was solution looking for problem. And the problem existed only for a handful of companies.
I think somehow, how cloud providers abstracted lot of these tools and databases, gave a better service and we kind of forgot about hadoop et al.
Overall, I agree with much of this post, but there are several caveats: 1) Mongo is a bad point of reference. The one lesson I've learned is that there is nothing Mongo does which postgresql doesn't do better. Big data solutions aren't nosql / mongo, but usually things like columnar databases, map/reduce, Cassandra, etc. 2) Plan for success 95% of businesses never become unicorns, but that's the goal for most (for th…
Sources ?
When I was hiring data scientists for a previous job, my favorite tricky question was "what stack/architecture would you build" with the somewhat detailed requirements of "6 TiB of data" in sight. I was careful not to require overly complicated sums, I simply said it's MAX 6TiB I patiently listened to all the big query hadoop habla-blabla, even asked questions about the financials (hardware/software/license BOM) and…
You have 6 TiB of ram?
But yeah, they might have that much RAM. At a rather small company I was at we had a third of it in the virtualisation cluster. I routinely put customer databases in the hundreds of gigabytes into RAM to do bug triage and fixing.
Overall, I agree with much of this post, but there are several caveats: 1) Mongo is a bad point of reference. The one lesson I've learned is that there is nothing Mongo does which postgresql doesn't do better. Big data solutions aren't nosql / mongo, but usually things like columnar databases, map/reduce, Cassandra, etc. 2) Plan for success 95% of businesses never become unicorns, but that's the goal for most (for th…
It's a second system syndrome + survivor bias thing I think: people who had to clean up the mess of a good MVP complaining about what wasn't done before. But the companies that DID do that planning and architecting before did not survive to be complained about.
When I was hiring data scientists for a previous job, my favorite tricky question was "what stack/architecture would you build" with the somewhat detailed requirements of "6 TiB of data" in sight. I was careful not to require overly complicated sums, I simply said it's MAX 6TiB I patiently listened to all the big query hadoop habla-blabla, even asked questions about the financials (hardware/software/license BOM) and…
To mitigate these bottlenecks you get fancy hardware (e.g oracle appliance) or you scale out (and get TCO/performance gains from separating storage and compute - which is how Snowflake sold 3x cheaper compared to appliances when they came out).
I believe that Trino on HDFS would be able to finish faster than awk on 6 enterprise disks for 6TB data.
In conclusion I would say that we should keep data local if possible but 6TB is getting into the realm where Big Data tech starts to be useful if you do it a lot.