Exploring DGRAPH


(Karan Kumar) #1

Hello Dgraph community. I was exploring Dgraph for a ancestry type of usecase where in I want to push my distributed file
system metadata and updates on to a backendb. Sample log line (path,size,user,timestamp,operation ) .
I want to visualize this on a weblayer where I can drill down starting from the root path.(based on size/approx size/child count) and find the large files or dir with high file count.
Also dashboard will show things like
(number of files created in the last 15 minutes by this user)
(Average size of files created in the last 15 minutes by this user) that will trigger alerts after they breach a certain threshold.

I come from the OLAP world so this question may sound obvious but can Dgraph support this usecase.
Number of Files would be approximately paths = 400 million
Fast query response times needed (2-3 seconds) for dashboard visualisation
ACID may not be required but updates are required. Eventual consistency works for me .


(Michel Conrado) #2

Yes

Which kind of updates?


(Karan Kumar) #3

Updates in the form of (path,size,user,timestamp,operation )

t1,a/b/c,1000 bytes, write
t2,a/b/c,1200 bytes, write

so in my web-layer, when I see a/b/c , I should see 1200 bytes.


(Michel Conrado) #4

I’m not sure if I understood it right. But Dgraph can’t do automatic updates based on some logic. This should be done in the application level(your app). But in Dgraph 1.1 you’ll be able to use upsert transactions. What makes any job like that easier.


(system) closed #5

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.