On Wed, Feb 22, 2012 at 16:19, Dimitry Sibiryakov <s...@ibphoenix.com> wrote: > 22.02.2012 15:13, ik wrote: >> And some data tasks require to act according to time >> rules. > > Such as?..
Queues that are specific for time, like Task Queues that can start for example at 10:00:25 pm but no longer then 10:05:40 pm that day . Let's say you have 5,000 tasks like that on queue. Each of them in different time scale and period. Cron is not the answer. Would you run your task every minute (what if the time to do it is now and you have 30 seconds to do it, you might miss that in cron that works on minutes based) ? would you keep it running as a daemon and collect this data with constantly selecting the proper records ? Would you create a copy of this in RAM to make access it faster, and actually create new table in memory that stores such records twice ? How would you go with that ? > >> At the moment, for doing this, you constantly require a daemon and not >> a cron based program to constantly access the database and change >> data. > > For example?.. > >> Wouldn't it be better if there was a better place to do >> it ? > > I don't see any better place for it. Certainly not in DBMS engine. But Oracle and MS, do think that their DBMS can provide such tasks. Even MySQL (that I don't really think that it's true DB) have this feature (but lack more important features). Let me ask you different questions: 1. Why do you think DBMS are not the proper engine for this ? 2. How would you recommend to do massive data manipulation on data in records in the most efficient way, that does not duplicate data, does not take too much memory from the machine, and does not duplicate your code for every project, but allow you to scale your tasks ? 3. So if DBMS should not have time based manipulation engine, then when you talk about DBMS, what exactly do you mean ? 4. Are you in the scholars that think that database engine's job is to store data, and make external applications to do the manipulation ? 5. What type of features (or direction towards features) do you think that database engine should have ? I can't understand the objection. The only think that I do see here are objections without explaining why. Why such type of features will be bad ? Why time based data manipulation are not good for databases but should be external ? Don't get me wrong, if I'll understand why, I will write that I was wrong, but I opened this for a discussion and the only thing I read here, is the word "No", and not "why not". > > -- > SY, SD. Ido > > ------------------------------------------------------------------------------ > Virtualization & Cloud Management Using Capacity Planning > Cloud computing makes use of virtualization - but cloud computing > also focuses on allowing computing to be delivered as a service. > http://www.accelacomm.com/jaw/sfnl/114/51521223/ > Firebird-Devel mailing list, web interface at > https://lists.sourceforge.net/lists/listinfo/firebird-devel ------------------------------------------------------------------------------ Virtualization & Cloud Management Using Capacity Planning Cloud computing makes use of virtualization - but cloud computing also focuses on allowing computing to be delivered as a service. http://www.accelacomm.com/jaw/sfnl/114/51521223/ Firebird-Devel mailing list, web interface at https://lists.sourceforge.net/lists/listinfo/firebird-devel
