Re: [DISCUSS] FLIP-8: Rescalable Non-Partitioned State
Hi Ufuk, Thanks for the great proposal I think this will be a very cool feature :) Option 1: I think the main problem here is sending all the state everywhere will not scale at all. I think this will even fail for some internal Flink operators (window timers I think are kept like this, maybe Im wrong here). The general problem here what we don't have with the key-value states is that the system can't do the repartitioning automatically. I think we should try to make abstractions that would allow the system to do this. Option 2: To be honest I don't completely get this approach, what do the indices mean in the get set methods? What happens if the same index is used from multiple operators? This may also suffers in scalability like option 1 (but as I said I dont get this completely :() I think another approach could be (might be similar what option 2 is trying to achieve) to provide a Set (or Map) like abstraction to keep the non partitioned states. Users could add/remove things from it at their on will, but the system would be free to redistribute the Sets between the operators. In practice this would mean for instance that the Kafka sources would store (partition, offset) tuples in the set but and every time in the open method they would check what is assigned to them (the system is free to decide). This of course would only work well if we can assume that distributing the states by equal numbers is desirable. Broadcast states: This might be a good time to think about broadcast states. Non-partitioned states that are the same at all subtasks, I think this comes up in a lot of use-cases (I know at least one myself haha) and it is pretty straight forward from a runtime perspective, the bigger question is the API. Cheers, Gyula Ufuk Celebi ezt írta (időpont: 2016. aug. 11., Cs, 19:20): > Hey all! > > I've created a short FLIP for rescalable non-partitioned state: > > > https://cwiki.apache.org/confluence/display/FLINK/FLIP-8%3A+Rescalable+Non-Partitioned+State > > This is related to an effort led by Till, Aljoscha, Stephan, and > Stefan to allow rescaling of Flink jobs (see FLINK-3755, FLINK-4381, > and > https://docs.google.com/document/d/1G1OS1z3xEBOrYD4wSu-LuBCyPUWyFd9l3T9WyssQ63w/edit > ). > > FLIP-8 provides some explanation about why we need rescaling of > non-partitioned state, how the APIs could look like and a high-level > implementation idea. > > I think the main point for discussion should be the APIs. I would like > to start working on an initial implementation next week. I would then > extend the implementation/proposed changes part after figuring out > more details/implications. > > Ufuk >
Re: [DISCUSS] FLIP-8: Rescalable Non-Partitioned State
Comments inline. On Thu, Aug 11, 2016 at 8:06 PM, Gyula Fóra wrote: > Option 1: > I think the main problem here is sending all the state everywhere will not > scale at all. I think this will even fail for some internal Flink operators > (window timers I think are kept like this, maybe Im wrong here). The > general problem here what we don't have with the key-value states is that > the system can't do the repartitioning automatically. I think we should try > to make abstractions that would allow the system to do this. The state size can definitely become a problem. For Kafka sources for example I don' think it would be problematic, but the timers it might be, yes. It definitely depends on the use case. In theory, we could also redistribute the list elements automatically, for example in a round robing fashion. The question is whether this will be enough in general. > > Option 2: > To be honest I don't completely get this approach, what do the indices mean > in the get set methods? What happens if the same index is used from > multiple operators? > This may also suffers in scalability like option 1 (but as I said I dont > get this completely :() Yes, I don't like it either. It's actually similar to Option 1 (from runtime perspective). I think the main question with Option 2 is whether we expose the API as an interface or a state class. If we go for this kind of interface we could parameterize the restore behaviour via the descriptor (e.g. flag to merge/union etc.). That should be more extensible than providing interfaces. > I think another approach could be (might be similar what option 2 is trying > to achieve) to provide a Set (or Map) like abstraction to keep the non > partitioned states. Users could add/remove things from it at their on will, > but the system would be free to redistribute the Sets between the > operators. In practice this would mean for instance that the Kafka sources > would store (partition, offset) tuples in the set but and every time in the > open method they would check what is assigned to them (the system is free > to decide). This of course would only work well if we can assume that > distributing the states by equal numbers is desirable. I think the same point applies to redistributing the list automatically (what I meant with whether it is "general enough"). I think what you describe here could be the list w/o unioning it. > > Broadcast states: > This might be a good time to think about broadcast states. Non-partitioned > states that are the same at all subtasks, I think this comes up in a lot of > use-cases (I know at least one myself haha) and it is pretty straight > forward from a runtime perspective, the bigger question is the API. Can you explain this a little more? Another open question (not addressed in the FLIP yet) is how we treat operators that have both keyed and non-keyed state. The current API kind of moves this question to the user.
Re: [DISCUSS] FLIP-8: Rescalable Non-Partitioned State
Hi, Let me try to explain what I mean by broadcast states. I think it is a very common pattern that people broadcast control messages to operators that also receive normal input events. some examples: broadcast a model for prediction, broadcast some information that should be the same at all subtasks but is evolving over time. At the same time these operators usually also do normal event processing based on the broadcasted input stream. There is currently no proper solution for this provided by the api. We can of course use connected operators or wrapper types and broadcast one of the input but there are several limitations. We cant use keyed states for instance becase that requires both inputs to be keyed (so we cant broadcast). Cheers, Gyula On Fri, Aug 12, 2016, 12:28 Ufuk Celebi wrote: > Comments inline. > > On Thu, Aug 11, 2016 at 8:06 PM, Gyula Fóra wrote: > > Option 1: > > I think the main problem here is sending all the state everywhere will > not > > scale at all. I think this will even fail for some internal Flink > operators > > (window timers I think are kept like this, maybe Im wrong here). The > > general problem here what we don't have with the key-value states is that > > the system can't do the repartitioning automatically. I think we should > try > > to make abstractions that would allow the system to do this. > > The state size can definitely become a problem. For Kafka sources for > example I don' think it would be problematic, but the timers it might > be, yes. It definitely depends on the use case. > > In theory, we could also redistribute the list elements automatically, > for example in a round robing fashion. The question is whether this > will be enough in general. > > > > > Option 2: > > To be honest I don't completely get this approach, what do the indices > mean > > in the get set methods? What happens if the same index is used from > > multiple operators? > > This may also suffers in scalability like option 1 (but as I said I dont > > get this completely :() > > Yes, I don't like it either. It's actually similar to Option 1 (from > runtime perspective). I think the main question with Option 2 is > whether we expose the API as an interface or a state class. If we go > for this kind of interface we could parameterize the restore behaviour > via the descriptor (e.g. flag to merge/union etc.). That should be > more extensible than providing interfaces. > > > I think another approach could be (might be similar what option 2 is > trying > > to achieve) to provide a Set (or Map) like abstraction to keep the > non > > partitioned states. Users could add/remove things from it at their on > will, > > but the system would be free to redistribute the Sets between the > > operators. In practice this would mean for instance that the Kafka > sources > > would store (partition, offset) tuples in the set but and every time in > the > > open method they would check what is assigned to them (the system is free > > to decide). This of course would only work well if we can assume that > > distributing the states by equal numbers is desirable. > > I think the same point applies to redistributing the list > automatically (what I meant with whether it is "general enough"). I > think what you describe here could be the list w/o unioning it. > > > > > Broadcast states: > > This might be a good time to think about broadcast states. > Non-partitioned > > states that are the same at all subtasks, I think this comes up in a lot > of > > use-cases (I know at least one myself haha) and it is pretty straight > > forward from a runtime perspective, the bigger question is the API. > > Can you explain this a little more? > > > > Another open question (not addressed in the FLIP yet) is how we treat > operators that have both keyed and non-keyed state. The current API > kind of moves this question to the user. >
Re: [DISCUSS] FLIP-8: Rescalable Non-Partitioned State
Hi Gyula, I was thinking about this as well, in the context of side-inputs, which would be a generalization of your use case. If I'm not mistaken. In my head I was calling it global state. Essentially, this state would be the same on all operators and when checkpointing you would only have to checkpoint the state of operator 0. Upon restore you would distribute this state to all operators again. Is this what you had in mind? Cheers, Aljoscha On Fri, 12 Aug 2016 at 13:07 Gyula Fóra wrote: > Hi, > Let me try to explain what I mean by broadcast states. > > I think it is a very common pattern that people broadcast control messages > to operators that also receive normal input events. > > some examples: broadcast a model for prediction, broadcast some information > that should be the same at all subtasks but is evolving over time. At the > same time these operators usually also do normal event processing based on > the broadcasted input stream. > > There is currently no proper solution for this provided by the api. We can > of course use connected operators or wrapper types and broadcast one of the > input but there are several limitations. We cant use keyed states for > instance becase that requires both inputs to be keyed (so we cant > broadcast). > > Cheers, > Gyula > > On Fri, Aug 12, 2016, 12:28 Ufuk Celebi wrote: > > > Comments inline. > > > > On Thu, Aug 11, 2016 at 8:06 PM, Gyula Fóra > wrote: > > > Option 1: > > > I think the main problem here is sending all the state everywhere will > > not > > > scale at all. I think this will even fail for some internal Flink > > operators > > > (window timers I think are kept like this, maybe Im wrong here). The > > > general problem here what we don't have with the key-value states is > that > > > the system can't do the repartitioning automatically. I think we should > > try > > > to make abstractions that would allow the system to do this. > > > > The state size can definitely become a problem. For Kafka sources for > > example I don' think it would be problematic, but the timers it might > > be, yes. It definitely depends on the use case. > > > > In theory, we could also redistribute the list elements automatically, > > for example in a round robing fashion. The question is whether this > > will be enough in general. > > > > > > > > Option 2: > > > To be honest I don't completely get this approach, what do the indices > > mean > > > in the get set methods? What happens if the same index is used from > > > multiple operators? > > > This may also suffers in scalability like option 1 (but as I said I > dont > > > get this completely :() > > > > Yes, I don't like it either. It's actually similar to Option 1 (from > > runtime perspective). I think the main question with Option 2 is > > whether we expose the API as an interface or a state class. If we go > > for this kind of interface we could parameterize the restore behaviour > > via the descriptor (e.g. flag to merge/union etc.). That should be > > more extensible than providing interfaces. > > > > > I think another approach could be (might be similar what option 2 is > > trying > > > to achieve) to provide a Set (or Map) like abstraction to keep the > > non > > > partitioned states. Users could add/remove things from it at their on > > will, > > > but the system would be free to redistribute the Sets between the > > > operators. In practice this would mean for instance that the Kafka > > sources > > > would store (partition, offset) tuples in the set but and every time in > > the > > > open method they would check what is assigned to them (the system is > free > > > to decide). This of course would only work well if we can assume that > > > distributing the states by equal numbers is desirable. > > > > I think the same point applies to redistributing the list > > automatically (what I meant with whether it is "general enough"). I > > think what you describe here could be the list w/o unioning it. > > > > > > > > Broadcast states: > > > This might be a good time to think about broadcast states. > > Non-partitioned > > > states that are the same at all subtasks, I think this comes up in a > lot > > of > > > use-cases (I know at least one myself haha) and it is pretty straight > > > forward from a runtime perspective, the bigger question is the API. > > > > Can you explain this a little more? > > > > > > > > Another open question (not addressed in the FLIP yet) is how we treat > > operators that have both keyed and non-keyed state. The current API > > kind of moves this question to the user. > > >
Re: [DISCUSS] FLIP-8: Rescalable Non-Partitioned State
Hi Aljoscha, Yes this is pretty much how I think about it as well. Basically the state in this case would be computed from the side inputs with the same state update logic on all operators. I think it is imprtant that operators compute their own state or at least observe all state changes otherwise a lot of things can get weird. Lets say for instance I am building a dynamic filter where new filter conditions are added /removed on the fly. For the sake of my argument lets also assume that initializing a new filter condition is a heavy operation. The global state in this case is the union of all filter conditions. If at any point in time the operators could only observe the current state we might end up with a very inefficient code, while if we observe all state changes individually (add 1 new filter) we can jus instantiate the new filter without worrying about the other ones. I am not completely sure if its clear what I am trying to say :D Gyula On Fri, Aug 12, 2016, 14:28 Aljoscha Krettek wrote: > Hi Gyula, > I was thinking about this as well, in the context of side-inputs, which > would be a generalization of your use case. If I'm not mistaken. In my head > I was calling it global state. Essentially, this state would be the same on > all operators and when checkpointing you would only have to checkpoint the > state of operator 0. Upon restore you would distribute this state to all > operators again. > > Is this what you had in mind? > > Cheers, > Aljoscha > > On Fri, 12 Aug 2016 at 13:07 Gyula Fóra wrote: > > > Hi, > > Let me try to explain what I mean by broadcast states. > > > > I think it is a very common pattern that people broadcast control > messages > > to operators that also receive normal input events. > > > > some examples: broadcast a model for prediction, broadcast some > information > > that should be the same at all subtasks but is evolving over time. At the > > same time these operators usually also do normal event processing based > on > > the broadcasted input stream. > > > > There is currently no proper solution for this provided by the api. We > can > > of course use connected operators or wrapper types and broadcast one of > the > > input but there are several limitations. We cant use keyed states for > > instance becase that requires both inputs to be keyed (so we cant > > broadcast). > > > > Cheers, > > Gyula > > > > On Fri, Aug 12, 2016, 12:28 Ufuk Celebi wrote: > > > > > Comments inline. > > > > > > On Thu, Aug 11, 2016 at 8:06 PM, Gyula Fóra > > wrote: > > > > Option 1: > > > > I think the main problem here is sending all the state everywhere > will > > > not > > > > scale at all. I think this will even fail for some internal Flink > > > operators > > > > (window timers I think are kept like this, maybe Im wrong here). The > > > > general problem here what we don't have with the key-value states is > > that > > > > the system can't do the repartitioning automatically. I think we > should > > > try > > > > to make abstractions that would allow the system to do this. > > > > > > The state size can definitely become a problem. For Kafka sources for > > > example I don' think it would be problematic, but the timers it might > > > be, yes. It definitely depends on the use case. > > > > > > In theory, we could also redistribute the list elements automatically, > > > for example in a round robing fashion. The question is whether this > > > will be enough in general. > > > > > > > > > > > Option 2: > > > > To be honest I don't completely get this approach, what do the > indices > > > mean > > > > in the get set methods? What happens if the same index is used from > > > > multiple operators? > > > > This may also suffers in scalability like option 1 (but as I said I > > dont > > > > get this completely :() > > > > > > Yes, I don't like it either. It's actually similar to Option 1 (from > > > runtime perspective). I think the main question with Option 2 is > > > whether we expose the API as an interface or a state class. If we go > > > for this kind of interface we could parameterize the restore behaviour > > > via the descriptor (e.g. flag to merge/union etc.). That should be > > > more extensible than providing interfaces. > > > > > > > I think another approach could be (might be similar what option 2 is > > > trying > > > > to achieve) to provide a Set (or Map) like abstraction to keep > the > > > non > > > > partitioned states. Users could add/remove things from it at their on > > > will, > > > > but the system would be free to redistribute the Sets between the > > > > operators. In practice this would mean for instance that the Kafka > > > sources > > > > would store (partition, offset) tuples in the set but and every time > in > > > the > > > > open method they would check what is assigned to them (the system is > > free > > > > to decide). This of course would only work well if we can assume that > > > > distributing the states by equal numbers is desirable. > > > > >
Re: [DISCUSS] FLIP-8: Rescalable Non-Partitioned State
I will update the design doc with more details for the Checkpointed variants and remove Option 2 (I think that's an orthogonal thing). The way I see it now, we should have base CheckpointedBase interface, have the current Checkpointed interface be a subclass for not repartitionable state. Then we have two other List-based variants: 1) Union List => on restore all state is unioned (what is currently in the design doc) 2) List => on restore state is automatically redistributed (if parallelism stays the same, state should go to the same sub tasks, but no guarantees when changed parallelism). Regarding the other thing you and Aljoscha discussed: I feel like that should be handled as part of the side input effort. Does that make sense? On Fri, Aug 12, 2016 at 3:11 PM, Gyula Fóra wrote: > Hi Aljoscha, > > Yes this is pretty much how I think about it as well. > > Basically the state in this case would be computed from the side inputs > with the same state update logic on all operators. I think it is imprtant > that operators compute their own state or at least observe all state > changes otherwise a lot of things can get weird. > > Lets say for instance I am building a dynamic filter where new filter > conditions are added /removed on the fly. For the sake of my argument lets > also assume that initializing a new filter condition is a heavy operation. > The global state in this case is the union of all filter conditions. > > If at any point in time the operators could only observe the current state > we might end up with a very inefficient code, while if we observe all state > changes individually (add 1 new filter) we can jus instantiate the new > filter without worrying about the other ones. > > I am not completely sure if its clear what I am trying to say :D > > Gyula > > On Fri, Aug 12, 2016, 14:28 Aljoscha Krettek wrote: > >> Hi Gyula, >> I was thinking about this as well, in the context of side-inputs, which >> would be a generalization of your use case. If I'm not mistaken. In my head >> I was calling it global state. Essentially, this state would be the same on >> all operators and when checkpointing you would only have to checkpoint the >> state of operator 0. Upon restore you would distribute this state to all >> operators again. >> >> Is this what you had in mind? >> >> Cheers, >> Aljoscha >> >> On Fri, 12 Aug 2016 at 13:07 Gyula Fóra wrote: >> >> > Hi, >> > Let me try to explain what I mean by broadcast states. >> > >> > I think it is a very common pattern that people broadcast control >> messages >> > to operators that also receive normal input events. >> > >> > some examples: broadcast a model for prediction, broadcast some >> information >> > that should be the same at all subtasks but is evolving over time. At the >> > same time these operators usually also do normal event processing based >> on >> > the broadcasted input stream. >> > >> > There is currently no proper solution for this provided by the api. We >> can >> > of course use connected operators or wrapper types and broadcast one of >> the >> > input but there are several limitations. We cant use keyed states for >> > instance becase that requires both inputs to be keyed (so we cant >> > broadcast). >> > >> > Cheers, >> > Gyula >> > >> > On Fri, Aug 12, 2016, 12:28 Ufuk Celebi wrote: >> > >> > > Comments inline. >> > > >> > > On Thu, Aug 11, 2016 at 8:06 PM, Gyula Fóra >> > wrote: >> > > > Option 1: >> > > > I think the main problem here is sending all the state everywhere >> will >> > > not >> > > > scale at all. I think this will even fail for some internal Flink >> > > operators >> > > > (window timers I think are kept like this, maybe Im wrong here). The >> > > > general problem here what we don't have with the key-value states is >> > that >> > > > the system can't do the repartitioning automatically. I think we >> should >> > > try >> > > > to make abstractions that would allow the system to do this. >> > > >> > > The state size can definitely become a problem. For Kafka sources for >> > > example I don' think it would be problematic, but the timers it might >> > > be, yes. It definitely depends on the use case. >> > > >> > > In theory, we could also redistribute the list elements automatically, >> > > for example in a round robing fashion. The question is whether this >> > > will be enough in general. >> > > >> > > > >> > > > Option 2: >> > > > To be honest I don't completely get this approach, what do the >> indices >> > > mean >> > > > in the get set methods? What happens if the same index is used from >> > > > multiple operators? >> > > > This may also suffers in scalability like option 1 (but as I said I >> > dont >> > > > get this completely :() >> > > >> > > Yes, I don't like it either. It's actually similar to Option 1 (from >> > > runtime perspective). I think the main question with Option 2 is >> > > whether we expose the API as an interface or a state class. If we go >> > > for this kind of interface we
Re: [DISCUSS] FLIP-8: Rescalable Non-Partitioned State
Hi, Yes I think it makes sense. :) Gyula On Fri, Aug 12, 2016, 17:02 Ufuk Celebi wrote: > I will update the design doc with more details for the Checkpointed > variants and remove Option 2 (I think that's an orthogonal thing). > > The way I see it now, we should have base CheckpointedBase interface, > have the current Checkpointed interface be a subclass for not > repartitionable state. Then we have two other List-based variants: > > 1) Union List => on restore all state is unioned (what is currently in > the design doc) > > 2) List => on restore state is automatically redistributed (if > parallelism stays the same, state should go to the same sub tasks, but > no guarantees when changed parallelism). > > > > Regarding the other thing you and Aljoscha discussed: I feel like that > should be handled as part of the side input effort. Does that make > sense? > > > > On Fri, Aug 12, 2016 at 3:11 PM, Gyula Fóra wrote: > > Hi Aljoscha, > > > > Yes this is pretty much how I think about it as well. > > > > Basically the state in this case would be computed from the side inputs > > with the same state update logic on all operators. I think it is imprtant > > that operators compute their own state or at least observe all state > > changes otherwise a lot of things can get weird. > > > > Lets say for instance I am building a dynamic filter where new filter > > conditions are added /removed on the fly. For the sake of my argument > lets > > also assume that initializing a new filter condition is a heavy > operation. > > The global state in this case is the union of all filter conditions. > > > > If at any point in time the operators could only observe the current > state > > we might end up with a very inefficient code, while if we observe all > state > > changes individually (add 1 new filter) we can jus instantiate the new > > filter without worrying about the other ones. > > > > I am not completely sure if its clear what I am trying to say :D > > > > Gyula > > > > On Fri, Aug 12, 2016, 14:28 Aljoscha Krettek > wrote: > > > >> Hi Gyula, > >> I was thinking about this as well, in the context of side-inputs, which > >> would be a generalization of your use case. If I'm not mistaken. In my > head > >> I was calling it global state. Essentially, this state would be the > same on > >> all operators and when checkpointing you would only have to checkpoint > the > >> state of operator 0. Upon restore you would distribute this state to all > >> operators again. > >> > >> Is this what you had in mind? > >> > >> Cheers, > >> Aljoscha > >> > >> On Fri, 12 Aug 2016 at 13:07 Gyula Fóra wrote: > >> > >> > Hi, > >> > Let me try to explain what I mean by broadcast states. > >> > > >> > I think it is a very common pattern that people broadcast control > >> messages > >> > to operators that also receive normal input events. > >> > > >> > some examples: broadcast a model for prediction, broadcast some > >> information > >> > that should be the same at all subtasks but is evolving over time. At > the > >> > same time these operators usually also do normal event processing > based > >> on > >> > the broadcasted input stream. > >> > > >> > There is currently no proper solution for this provided by the api. We > >> can > >> > of course use connected operators or wrapper types and broadcast one > of > >> the > >> > input but there are several limitations. We cant use keyed states for > >> > instance becase that requires both inputs to be keyed (so we cant > >> > broadcast). > >> > > >> > Cheers, > >> > Gyula > >> > > >> > On Fri, Aug 12, 2016, 12:28 Ufuk Celebi wrote: > >> > > >> > > Comments inline. > >> > > > >> > > On Thu, Aug 11, 2016 at 8:06 PM, Gyula Fóra > >> > wrote: > >> > > > Option 1: > >> > > > I think the main problem here is sending all the state everywhere > >> will > >> > > not > >> > > > scale at all. I think this will even fail for some internal Flink > >> > > operators > >> > > > (window timers I think are kept like this, maybe Im wrong here). > The > >> > > > general problem here what we don't have with the key-value states > is > >> > that > >> > > > the system can't do the repartitioning automatically. I think we > >> should > >> > > try > >> > > > to make abstractions that would allow the system to do this. > >> > > > >> > > The state size can definitely become a problem. For Kafka sources > for > >> > > example I don' think it would be problematic, but the timers it > might > >> > > be, yes. It definitely depends on the use case. > >> > > > >> > > In theory, we could also redistribute the list elements > automatically, > >> > > for example in a round robing fashion. The question is whether this > >> > > will be enough in general. > >> > > > >> > > > > >> > > > Option 2: > >> > > > To be honest I don't completely get this approach, what do the > >> indices > >> > > mean > >> > > > in the get set methods? What happens if the same index is used > from > >> > > > multiple operators? > >> > > > This may als
Re: [DISCUSS] FLIP-8: Rescalable Non-Partitioned State
Hi, I think we have to rethink a bit how the state interfaces work. Having an interface where you snapshot a type T and restore a type T are not well suited for dealing with job updates/state updates. For example, let's look at the current Checkpointed interface. The value that you return is serialized using Java Serialization. This means, that you can never change the fields of that type T, i.e. you can never update your data type if you want to restore from a savepoint of an earlier version. The only workaround to this is to change Checkpointed to Checkpointed and manually check the type of the object that you get in restore and cast it. In the long run, this will lead to everyone implementing Checkpointed and doing manual (non type safe) casts. Cheers, Aljoscha On Sun, 14 Aug 2016 at 06:54 Gyula Fóra wrote: > Hi, > Yes I think it makes sense. :) > > Gyula > > On Fri, Aug 12, 2016, 17:02 Ufuk Celebi wrote: > > > I will update the design doc with more details for the Checkpointed > > variants and remove Option 2 (I think that's an orthogonal thing). > > > > The way I see it now, we should have base CheckpointedBase interface, > > have the current Checkpointed interface be a subclass for not > > repartitionable state. Then we have two other List-based variants: > > > > 1) Union List => on restore all state is unioned (what is currently in > > the design doc) > > > > 2) List => on restore state is automatically redistributed (if > > parallelism stays the same, state should go to the same sub tasks, but > > no guarantees when changed parallelism). > > > > > > > > Regarding the other thing you and Aljoscha discussed: I feel like that > > should be handled as part of the side input effort. Does that make > > sense? > > > > > > > > On Fri, Aug 12, 2016 at 3:11 PM, Gyula Fóra > wrote: > > > Hi Aljoscha, > > > > > > Yes this is pretty much how I think about it as well. > > > > > > Basically the state in this case would be computed from the side inputs > > > with the same state update logic on all operators. I think it is > imprtant > > > that operators compute their own state or at least observe all state > > > changes otherwise a lot of things can get weird. > > > > > > Lets say for instance I am building a dynamic filter where new filter > > > conditions are added /removed on the fly. For the sake of my argument > > lets > > > also assume that initializing a new filter condition is a heavy > > operation. > > > The global state in this case is the union of all filter conditions. > > > > > > If at any point in time the operators could only observe the current > > state > > > we might end up with a very inefficient code, while if we observe all > > state > > > changes individually (add 1 new filter) we can jus instantiate the new > > > filter without worrying about the other ones. > > > > > > I am not completely sure if its clear what I am trying to say :D > > > > > > Gyula > > > > > > On Fri, Aug 12, 2016, 14:28 Aljoscha Krettek > > wrote: > > > > > >> Hi Gyula, > > >> I was thinking about this as well, in the context of side-inputs, > which > > >> would be a generalization of your use case. If I'm not mistaken. In my > > head > > >> I was calling it global state. Essentially, this state would be the > > same on > > >> all operators and when checkpointing you would only have to checkpoint > > the > > >> state of operator 0. Upon restore you would distribute this state to > all > > >> operators again. > > >> > > >> Is this what you had in mind? > > >> > > >> Cheers, > > >> Aljoscha > > >> > > >> On Fri, 12 Aug 2016 at 13:07 Gyula Fóra wrote: > > >> > > >> > Hi, > > >> > Let me try to explain what I mean by broadcast states. > > >> > > > >> > I think it is a very common pattern that people broadcast control > > >> messages > > >> > to operators that also receive normal input events. > > >> > > > >> > some examples: broadcast a model for prediction, broadcast some > > >> information > > >> > that should be the same at all subtasks but is evolving over time. > At > > the > > >> > same time these operators usually also do normal event processing > > based > > >> on > > >> > the broadcasted input stream. > > >> > > > >> > There is currently no proper solution for this provided by the api. > We > > >> can > > >> > of course use connected operators or wrapper types and broadcast one > > of > > >> the > > >> > input but there are several limitations. We cant use keyed states > for > > >> > instance becase that requires both inputs to be keyed (so we cant > > >> > broadcast). > > >> > > > >> > Cheers, > > >> > Gyula > > >> > > > >> > On Fri, Aug 12, 2016, 12:28 Ufuk Celebi wrote: > > >> > > > >> > > Comments inline. > > >> > > > > >> > > On Thu, Aug 11, 2016 at 8:06 PM, Gyula Fóra > > > >> > wrote: > > >> > > > Option 1: > > >> > > > I think the main problem here is sending all the state > everywhere > > >> will > > >> > > not > > >> > > > scale at all. I think this will even fail for some internal > Flink
