Re: Janusgraph Hadoop Spark standalone cluster - Janusgraph job always creates constant number 513 of Spark tasks
Varun Ganesh <operatio...@...>
Hi,toggle quoted message Show quoted text
I am facing this same issue. I am using SparkGraphComputer to read from Janusgraph backed by cassandra. `g.V().count()` takes about 3 minutes to load just two rows that I have in the graph.
I see that about 257 tasks are created. In my case, I am seeing parallelism in the spark cluster that I am using but each task seems to take about ~5 seconds on average and there is no obvious reason why.
(I can attach a page from the Spark UI and also the properties file I am using, but I am unable to find the option to)
Would appreciate any input on solving this. Thank you!
On Friday, December 6, 2019 at 5:04:55 AM UTC-5 s...@... wrote: