Re: List State in RichWindowFunction leads to RocksDb memory leak

2017-02-27 Thread Aljoscha Krettek
gt; context? > > > > Seth Wiesman > > > > *From: *Seth Wiesman > *Reply-To: *"user@flink.apache.org" > *Date: *Friday, February 24, 2017 at 2:51 PM > > > *To: *"user@flink.apache.org" > *Subject: *Re: List State in RichWindowFunction

Re: List State in RichWindowFunction leads to RocksDb memory leak

2017-02-24 Thread Seth Wiesman
tek Reply-To: "user@flink.apache.org" Date: Friday, February 24, 2017 at 12:09 PM To: "user@flink.apache.org" Subject: Re: List State in RichWindowFunction leads to RocksDb memory leak Hi Seth, yes, this is a thorny problem but I actually see one additional possible solution (

Re: List State in RichWindowFunction leads to RocksDb memory leak

2017-02-24 Thread Seth Wiesman
;user@flink.apache.org" Date: Friday, February 24, 2017 at 12:09 PM To: "user@flink.apache.org" Subject: Re: List State in RichWindowFunction leads to RocksDb memory leak Hi Seth, yes, this is a thorny problem but I actually see one additional possible solution (that will, however, brea

Re: List State in RichWindowFunction leads to RocksDb memory leak

2017-02-24 Thread Aljoscha Krettek
Hi Seth, yes, this is a thorny problem but I actually see one additional possible solution (that will, however, break other possible use cases. First, regarding your solution 1): We are working on adding this for ProcessWindowFunction: https://issues.apache.org/jira/browse/FLINK-4953. ProcessWindo

List State in RichWindowFunction leads to RocksDb memory leak

2017-02-23 Thread Seth Wiesman
I am working on a program that uses a complex window and have run into some issues. It is a 1 hour window with 7 days allowed lateness including a custom trigger that gives us intermediate results every 5 minutes of processing time until the end of 7 days event time when a final fire is triggere