Skip to content

[SPARK-5282][mllib]: RowMatrix easily gets int overflow in the memory size warning #4069

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
hhbyyh wants to merge 2 commits into from
Closed

[SPARK-5282][mllib]: RowMatrix easily gets int overflow in the memory size warning #4069

Changes from 1 commit
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
7afac23
5282: fix int overflow in the warning
hhbyyh Jan 17, 2015
e54e5c8
change to MB based number
hhbyyh Jan 18, 2015
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion mllib/src/main/scala/org/apache/spark/mllib/linalg/distributed/RowMatrix.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -131,7 +131,7 @@ class RowMatrix(
throw new IllegalArgumentException(s"Argument with more than 65535 cols: $cols")
}
if (cols > 10000) {
val mem = cols * cols * 8
val mem = (math.pow(cols, 2) * 8).formatted("%,.0f")
Copy link
Member

@srowen srowen Jan 16, 2015

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I know it's minor, but I think it might be simpler to report megabytes in the warning message, since the reported size is at least 800MB. Then this can be simply val memMB = (cols.toLong * cols) / 125000; that multiplication can't overflow.

Copy link
Contributor Author

@hhbyyh hhbyyh Jan 16, 2015

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Surely I can do it. I was worrying some one may ask for the accurate number in the PR~~ :)

Another thing is, /125000 might should be changed to >> 15?

Copy link
Member

@srowen srowen Jan 16, 2015

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

>> 17 ? a mebibyte is 2^20 bytes so that would make sense. I don't know what the overall stance on mega vs mebibyte is in the nomenclature here but think everything is megabytes, in which case / 125000 is right.

Copy link
Contributor Author

@hhbyyh hhbyyh Jan 16, 2015

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

yes, 17...I'll send update tomorrow morning.

Copy link
Contributor Author

@hhbyyh hhbyyh Jan 16, 2015

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Maybe this 1000/1024 thing is why they send actual bytes in the beginning... :)

logWarning(s"$cols columns will require at least $mem bytes of memory!")
}
}
Expand Down