Hi, I'm trying to see how "Navigator" product looks and operates. Is the 'Audit' capability of the QuickStart VM part of 'Navigator', or is the Navigator auditing feature different from that?
I understand that there was an error on the website saying Navigator was in the QuickStart VM, but in fact is only in the Enterprise CDH4 download ... I have been unable to install that on two supported OS's, and haven't had a response the past week to that post, which is:
So, I'm wondering if there actually is a difference between "Navigator" and the QuickStart 'Audit' features.
Short answer: Navigator is not part of the QuickStart VM.
To try out Navigator and get a taste of what Hadoop auditing is like, follow this http://www.cloudera.com/content/cloudera-content/cloudera-docs/CM4Ent/latest/Cloudera-Manager-Instal.... Right after you log into that CM, you'll see a page that lets you pick from Standard, Trial and Enterprise. Pick Trial.
The "Audits" page you see in CM shows you audits from 2 sources:
1. Admistrative actions in CM, e.g. config changes, hdfs restart, new host added, etc.
2. User actions in the cluster, e.g. somebody touched a file in HDFS, read from a table via a Hive/Impala query, etc.
#2 is only shown if you have Navigator. Otherwise, you're just looking at #1, which is always available.
The link you gave is installing the system onto EC2 ...
That's quite a lot of work, which I've already tried: I've installed the Enterprise CDH4 system onto two supported platforms,
Ubuntu 12.04 64bit and CentOS 6.4 64 bit, and each (of many) times the 'heatbeat' failed to work due ostensibly to port conflicts.... which is difficult to understand since I could see they were available, and twice I attempted to install on a clean VM.
Is there a screen shot of Navigator I can see? I really simply want to get an idea how it works.
Here are a couple of screenshots, showing:
- The audits UI of a mix of administrative audits and user activity audits.
- The filtering and drill down of the audits.
- What you'd get from the API, which has more information than the UI, e.g. the raw query text. The API also allows you to do filtering.
I was suggesting your install on ec2, because that eliminates any local networking issues that you may have. It's the easiest way to set up a CM + CDH for testing things out. And it costs less than a burrito.
Ah, that's terrific, thank you ... those screen shots are exactly what I wanted, viz. a taste of how it looks and works, thank you.
I didn't want to try the EC2 install, since I'd tried on two of the explicitly supported platforms already, including a clean VM, and failed, at the heartbeat stage after the install.
Thank you again!