Data Analytics Studio (DAS)
DAS is a memory-heavy and a disk-light application. For optimum performance, consider profiling the CPU cores, memory allocation, and disk space depending upon the number of users, the total number of databases and tables, and the number of queries in the system.
If you are setting up a high-availability cluster, then add additional cores and memory for the load balancer.
The following table provides component-wise recommendation for provisioning CPU, memory, and disk space. These recommendations are approximated considering 10 users, 10,000 Hive tables, 100 parallel Event Processor threads, and 40,000 queries.
| DAS component | CPU | Memory | Local Disk |
|---|---|---|---|
| Webapp |
*The number of cores that you allocate need to be proportional to U. |
*The amount of memory that you allocate need to be proportional to U and T. |
*The amount of disk space that you allocate need to be proportional to U. |
| Event Processor |
*The number of cores that you allocate need to be proportional to P. |
*The amount of memory that you allocate need to be proportional to P and T. |
*The disk space is primarily used for logs, and can remain constant. |
| Database |
*The number of cores that you allocate need to be proportional to (P + U). |
*The amount of memory that you allocate need to be proportional to (T + Q). |
*The amount of disk space that you allocate need to be proportional to (T + U + Q). |
Where,
U is the number of users concurrently accessing the DAS Webapp
T is the number of tables in Hive
P denotes the parallelism configured in the DAS Event Processor
Q is the total number of queries in the system
| Default Port Number | Description |
|---|---|
| 30900 | Event Processor server port |
| 30901 | Event Processor admin server port |
| 30800 | Webapp server port |
| 30801 | Webapp admin port |
