Sponsored Content
Full Discussion: Heap fragementation on HPUX
Operating Systems HP-UX Heap fragementation on HPUX Post 302330107 by atgoel on Tuesday 30th of June 2009 08:48:19 AM
Old 06-30-2009
Heap fragementation on HPUX

The Resident size(as observed from top) of my process is increasing. But, the behaviour is very random. My process works on request reponse model. So when i put some request load on my process the memory starts increasing. For initial few hours (approx ~3 hrs) it increase at a rapid rate and after that increase become stable. Some time after 30 min a 16 kb increase is observed some time after 2-3 hrs. And some times an increase of even 128kb is observed. Behaviour is really random. Attached are the some graph of resident memory of process taken at interval of 10 min each.

I tried wdb in BATCH mode with my process. But, it didnot detected any leaks. It produced only heap profile files.

Is there any way to find out that the wt=hether increase is due to fragmentation? Because lot of malloc are happening in my process.
Platform is HP-UX B.11.23 on HP itanium server. I can use wdb and glance tool(I only have working knowledge of both of these Smilie )

Also, considering the fact that memory increase is after 2-3 hr run how can i work with wdb in debugger mode. Means is it possible to automate printing the heap information and continuing from that point in gdb/wdb automatically?
 

9 More Discussions You Might Find Interesting

1. Filesystems, Disks and Memory

heap size!

I'm a new guy to this field and I'm learning a lot about UNIX! Can any explan to me what exactly does 'heap size' mean and how can i increase the size for AIX 4.3.3? (2 Replies)
Discussion started by: i2admin
2 Replies

2. Filesystems, Disks and Memory

heap size for JVM!

Hi all, Thanks 'thehoghunter' and 'hugo' for the comments! I've to increase the size of the heap size for AIX 4.3.3. Now what's the command that I've and also is it something similar to growing the file system in Solaris (growfs) (1 Reply)
Discussion started by: i2admin
1 Replies

3. UNIX for Advanced & Expert Users

monitor jvm heap size

I'm running websphere 4.5 on AIX 5 with java 1.3 and would like to find out the following: How much memory is allocated to each JVM, and how much of the allocated heap size is actually being used by a specific JVM? (0 Replies)
Discussion started by: rein
0 Replies

4. AIX

AIX 5.3 Heap Memory with SAP

Any idea whether this parameter can be set / adjusted and where please? (3 Replies)
Discussion started by: johnf
3 Replies

5. Programming

Heap and stack

Hi, I have a basic doubt here. Consider the following code snippet: main() { int *a; . . } Here the memory for a gets allocated in heap or stack. (5 Replies)
Discussion started by: naan
5 Replies

6. Shell Programming and Scripting

Need Script to Use CPUs on a HPUX server to simulate Workload Manager on HPUX.

I am running HPUX and using WLM (workload manager). I want to write a script to fork CPUs to basically take CPUs from other servers to show that the communication is working and CPU licensing is working. Basically, I want to build a script that will use up CPU on a server. Any ideas? (2 Replies)
Discussion started by: cpolikowsky
2 Replies

7. HP-UX

Heap fragmentation on HPUX

Hi All, We are facing issues on HPUX with the C heap region growing. We use a product for CRM by name ClarifyCRM and it uses a native layer for DB access. so there are best practices in place to actual control memory. recently we have seen issues that the C heap region is growing faster than... (0 Replies)
Discussion started by: ramchand75
0 Replies

8. HP-UX

pwage-hpux-T for Trusted HPUX servers

I'm sharing this in case anybody needs it. Modified from the original solaris pwage script. This modified hpux script will check /etc/password file on hpux trusted systems search /tcb and grep the required u_succhg field. Calculate days to expiry and notify users via email. original solaris... (2 Replies)
Discussion started by: sparcguy
2 Replies

9. Programming

find size of heap allocated

I want to find the size of the total memory allocated on the heap for the following statement: int* a = new int;How can I use the sizeof operator for this? I used: printf("\t===> %d\n",sizeof(*a)); Is this statement correct? I have asked the question because when I checked the memory of... (13 Replies)
Discussion started by: rupeshkp728
13 Replies
iotop(1m)							   USER COMMANDS							 iotop(1m)

NAME
iotop - display top disk I/O events by process. Uses DTrace. SYNOPSIS
iotop [-C] [-D|-o|-P] [-j|-Z] [-d device] [-f filename] [-m mount_point] [-t top] [interval [count]] DESCRIPTION
iotop tracks disk I/O by process, and prints a summary report that is refreshed every interval. This is measuring disk events that have made it past system caches. Since this uses DTrace, only users with root privileges can run this command. OPTIONS
-C don't clear the screen -D print delta times - elapsed, us -j print project ID -o print disk delta times, us -P print %I/O (disk delta times) -Z print zone ID -d device instance name to snoop (eg, dad0) -f filename full pathname of file to snoop -m mount_point mountpoint for filesystem to snoop -t top print top number only EXAMPLES
Default output, print summary every 5 seconds # iotop One second samples, # iotop 1 print %I/O (time based), # iotop -P Snoop events on the root filesystem only, # iotop -m / Print top 20 lines only, # iotop -t 20 Print 12 x 5 second samples, scrolling, # iotop -C 5 12 FIELDS
UID user ID PID process ID PPID parent process ID PROJ project ID ZONE zone ID CMD command name for the process DEVICE device name MAJ device major number MIN device minor number D direction, Read or Write BYTES total size of operations, bytes ELAPSED total elapsed times from request to completion, us (this is the elapsed time from the disk request (strategy) to the disk completion (iodone)) DISKTIME total times for disk to complete request, us (this is the time for the disk to complete that event since it's last event (time between iodones), or, the time to the strategy if the disk had been idle) %I/O percent disk I/O, based on time (DISKTIME) load 1 minute load average disk_r total disk read Kb for sample disk_w total disk write Kb for sample DOCUMENTATION
See the DTraceToolkit for further documentation under the Docs directory. The DTraceToolkit docs may include full worked examples with ver- bose descriptions explaining the output. EXIT
iotop will run forever until Ctrl-C is hit, or the specified interval is reached. AUTHOR
Brendan Gregg [Sydney, Australia] SEE ALSO
iosnoop(1M), dtrace(1M) version 0.75 Oct 25, 2005 iotop(1m)
All times are GMT -4. The time now is 05:42 PM.
Unix & Linux Forums Content Copyright 1993-2022. All Rights Reserved.
Privacy Policy