in reply to Re^3: out of memory problem
in thread out of memory problem
perl -v This is perl, v5.8.8 built for i486-linux-gnu-thread-multi Copyright 1987-2006, Larry Wall
|
|---|
| Replies are listed 'Best First'. | |
|---|---|
|
Re^5: out of memory problem
by BrowserUk (Patriarch) on Sep 13, 2008 at 11:13 UTC | |
Try again: I requested perl -V. Note: Uppercase 'V'. And don't keep truncating the output if you want a resolution. Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error.
"Science is about questioning the status quo. Questioning authority".
In the absence of evidence, opinion is indistinguishable from prejudice.
| [reply] [d/l] |
by lightoverhead (Pilgrim) on Sep 13, 2008 at 20:11 UTC | |
| [reply] [d/l] |
by BrowserUk (Patriarch) on Sep 13, 2008 at 20:58 UTC | |
Amongst others, there is one very strong clue that your perl is limited to 4GB: ptrsize=4,. With 4-byte pointers 2**32/4GB is the maximum memory the process will be able to address. But, a good chunk of that address space will be devoted to the OS. Often the top 1GB. So if top was showing you that 40% (3.2GB) of memory was used, with some of that occupied by other processes, system drivers etc., the realistic expectation is that your Perl programs will be limited to 3GB. As for how to resolve the problem, the pat answer would be "use less memory", but that's probably not very helpful. Some possibilities are: build a 64-bit perl; split the data into 2 processes and have the code that uses it do so from a 3rd process via sockets or pipes; store the data more efficiently. It is often possible to reduce the memory usage, but to suggest ways you'll have to give us more information. Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error.
"Science is about questioning the status quo. Questioning authority".
In the absence of evidence, opinion is indistinguishable from prejudice.
| [reply] [d/l] [select] |
by blazar (Canon) on Sep 14, 2008 at 11:01 UTC | |
And don't keep truncating the output if you want a resolution I personally believe that while this suggestion of your is certainly sensible, and the most important one as far as the target of solving his problem is concerned, it would be also recommendable to tell the OP to wrap the output up in <readmore> tags for viewing clarity in the various sections listings. | [reply] [d/l] [select] |
by BrowserUk (Patriarch) on Sep 14, 2008 at 11:57 UTC | |
o wrap the output up in | [reply] |
by blazar (Canon) on Sep 14, 2008 at 12:51 UTC | |