0

A counterpart to memory leak, but involving CPU time instead - the program's CPU usage slowly climbing to 100% and then the whole system slowing down - without memory footprint changing.

An example where this could happen - a program aggregating old data, polling the queue for samples in given interval; while the database purges obsolete data, the program checks for data since the beginning of operation, picking increasingly growing number of "no record" as the purged period grows.

Or a program, where the complexity of calculation grows as O(exp(n)), while n is proportional to the data set size; the data set grows normally, never exceeding foreseen, allocated space, but the algorithm processing speed grinds to a halt.

Is there an official name for this kind of problem/bug?

closed as primarily opinion-based by gnat, Jörg W Mittag, Ixrec, Blrfl, Bart van Ingen Schenau Apr 9 '16 at 18:10

Many good questions generate some degree of opinion based on expert experience, but answers to this question will tend to be almost entirely based on opinions, rather than facts, references, or specific expertise. If this question can be reworded to fit the rules in the help center, please edit the question.

  • No, I would speak of useless computation, or waste of CPU time. – Basile Starynkevitch Apr 9 '16 at 12:09
  • @BasileStarynkevitch: Still, that waste may be constant, a problem of optimality but not one that makes for a critical bug, one that makes the program useless in the long run. Imagine you have to quit and restart the app every hour or two to keep it running. – SF. Apr 9 '16 at 12:45
  • I'm seeing three close votes for "primarily OPINION based" on a question that asks for an OFFICIAL name for given effect. WTF?? Do you all really think calling the situation where a program repeatedly doesn't deallocate memory where it should, "Memory Leak" is also an opinion-based issue? – SF. Apr 9 '16 at 14:29
  • I don't think you'll find an open question here that asks if there's a name for allocating memory and not freeing it. – Blrfl Apr 9 '16 at 15:26
  • 1
    One possible cause would be a callback leak : when an operation takes longer and longer, it can be because the data being processed becomes more complex and takes more time (but that isn't happening here because memory usage is constant), or because more and more logic is being added to the handler. If the data complexity IS growing, but computation is growing more quickly, that's poor scalability as Erik already commented. – Ben Voigt Apr 10 '16 at 1:55
2

Is there an official name for this kind of problem/bug?

No there isn't. (This is a fact, not an opinion.)

There isn't an unofficial one either.

  • CPU Load Problems. Thats the name. – rxantos Apr 10 '16 at 1:00
  • Or "a bug". But neither of these captures the specific characteristic in the way that "memory leak" does. Hence my answer. – Stephen C Apr 10 '16 at 2:20
2

A system taking an inordinate time to compute something is usually just presumed to hang or be hung. The idea is that a computation that takes exponential time quickly becomes so slow that you can't distinguish it from an actual infinite loop, e.g. a total hang.

Not the answer you're looking for? Browse other questions tagged or ask your own question.