Home/ Questions/Q 6796033
In Process

The Archive Base Latest Questions

Editorial Team
  • 0
Asked: 2026-05-26T18:25:05+00:00

I wrote a small geoip lookup linux daemon in C++, and I’m getting two

  • 0

I wrote a small geoip lookup linux daemon in C++, and I’m getting two different results and I’m wondering why.

This is the source code: https://github.com/homer6/geoipd
I’ve checked it with valgrind for memory leaks and there weren’t any.

I have two web servers, both running the same Ubuntu image on Amazon EC2 and both behind a load balancer that is ensuring they each get about the same amount of traffic. I installed the daemon on both servers and I’ve let them run in production for a few days now.

Everything is working as expected, but the output from “ps aux” is slightly different for the two instances.

Server 1:

USER       PID %CPU %MEM    VSZ   RSS TTY      STAT START   TIME
1004     28889  0.0  6.7 640288 517692 ?       Ss   Nov09   0:03

Server 2:

USER       PID %CPU %MEM    VSZ   RSS TTY      STAT START   TIME
1004     16587  0.0  6.7 574752 517688 ?       Ss   Nov09   0:02

My question is two-parted:

  1. Why is there such a big difference in the virtual memory usage?
  2. Why is there a difference in the resident memory usage?

Thanks in advance…

UPDATE:
I remember checking the VM size when I first launched the processes. They were both at 574752 (or slightly under 575 MB) and they both had the same value.

Share

Leave an answer

You must login to add an answer.

Need An Account,

1 Answer

  1. Editorial Team

    From these 2 extra lines in server 1’s maps:

    7f003ed31000-7f003ed32000 ---p 00000000 00:00 0
7f003ed32000-7f003f532000 rw-p 00000000 00:00 0

    it seems server 1 has allocated (but not initializing yet) memory for some object that server 2 has not yet. This is not necessarily bad; each kernel is most likely handling the memory properly, just different.

    You can run size geoip_server.o to make sure that the bss (uninitialized static data) is the same for the 2 daemons on each system:

    [ 18:10 root@hozbox / ]# size geoip_server.o
   text    data    bss         dec         hex       filename
   77727   1088772 200003752   201170251   bfd9d4b   geoip_server.o

    these numbers are made up as I do not have cmake 2.8.2 so I was unable to compile your code

    Additionally, you can cat /proc/meminfo on each system to see the specifics of how each kernel is managing its memory:

    [ 18:10 root@hozbox / ]# cat /proc/meminfo
MemTotal:         443880 kB
MemFree:            7792 kB
Buffers:             788 kB
Cached:            24468 kB
SwapCached:        87512 kB
Active:           189460 kB
Inactive:         193316 kB
Active(anon):     178604 kB
Inactive(anon):   178956 kB
Active(file):      10856 kB
Inactive(file):    14360 kB
Unevictable:           0 kB
Mlocked:               0 kB
HighTotal:             0 kB
HighFree:              0 kB
LowTotal:         443880 kB
LowFree:            7792 kB
SwapTotal:       4816892 kB
SwapFree:        3856464 kB
Dirty:                12 kB
Writeback:             0 kB
AnonPages:        305332 kB
Mapped:             7812 kB
Shmem:                60 kB
Slab:              17628 kB
SReclaimable:       8712 kB
SUnreclaim:         8916 kB
KernelStack:        1320 kB
PageTables:         4592 kB
NFS_Unstable:          0 kB
Bounce:                0 kB
WritebackTmp:          0 kB
CommitLimit:     5038832 kB
Committed_AS:    1910872 kB
VmallocTotal:     570424 kB
VmallocUsed:       32424 kB
VmallocChunk:     524124 kB
HardwareCorrupted:     0 kB
HugePages_Total:       0
HugePages_Free:        0
HugePages_Rsvd:        0
HugePages_Surp:        0
Hugepagesize:       4096 kB
DirectMap4k:       19392 kB
DirectMap4M:      438272 kB

    This research paper from the University of Alberta on Understanding Memory is a very good read and may help explain the why a difference in VSZ for the same process running on 2 different systems can be normal.

    Some references and resources:

    procfs wiki page
    Runtime Memory Measurement
    Linux Kernel Documentation on /proc
    https://unix.stackexchange.com/questions/6301/how-do-i-read-from-proc-pid-mem-under-linux

    man proc:

    /proc/[pid]/maps
    A file containing the currently mapped memory regions
    and their access permissions. The format is:

    address           perms offset  dev   inode   pathname
08048000-08056000 r-xp 00000000 03:0c 64593   /usr/sbin/gpm
08056000-08058000 rw-p 0000d000 03:0c 64593   /usr/sbin/gpm
08058000-0805b000 rwxp 00000000 00:00 0 
40000000-40013000 r-xp 00000000 03:0c 4165    /lib/ld-2.2.4.so 
40013000-40015000 rw-p 00012000 03:0c 4165    /lib/ld-2.2.4.so
4001f000-40135000 r-xp 00000000 03:0c 45494   /lib/libc-2.2.4.so
40135000-4013e000 rw-p 00115000 03:0c 45494   /lib/libc-2.2.4.so
4013e000-40142000 rw-p 00000000 00:00 0 
bffff000-c0000000 rwxp 00000000 00:00 0

    where “address” is the address space in the process that it occupies,
    “perms” is a set of permissions,
    “offset” is the offset into the file/whatever, “dev” is the
    device (major:minor), and “inode” is the
    inode on that device. 0 indicates that no inode is associated with the
    memory region, as the case would be with BSS (uninitialized data).

    • 0