pt-summary Memory: Total reports M instead of G

[Michael Coburn]

It appears that the Memory section of pt-summary is incorrectly reporting total memory. In my case I have an Ubuntu VirtualBox guest with 1024MB RAM so it should say Total 1G but in the Memory section, Total line it reports 1M

# Memory #####################################################
       Total | 1.0M
        Free | 81.3M
        Used | physical = 920.8M, swap allocated = 510.0M, swap used = 41.7M, virtual = 962.5M

root@michael-VirtualBox:~# free
             total used free shared buffers cached
Mem: 1026080 955496 70584 0 146136 500328
-/+ buffers/cache: 309032 717048
Swap: 522236 42392 479844

root@michael-VirtualBox:~# uname -a
Linux michael-VirtualBox 3.0.0-12-generic #20-Ubuntu SMP Fri Oct 7 14:50:42 UTC 2011 i686 i686 i386 GNU/Linux
root@michael-VirtualBox:~# tail -1 /etc/apt/sources.list
deb-src http://repo.percona.com/apt oneiric main

root@michael-VirtualBox:~# pt-summary --version
local: 1: 2.1.1: bad variable name
pt-summary

Full command output from pt-summary:

root@michael-VirtualBox:~# PTDEBUG=1; pt-summary
# Percona Toolkit System Summary Report ######################
        Date | 2012-05-02 17:54:59 UTC (local TZ: EDT -0400)
    Hostname | michael-VirtualBox
      Uptime | 5:53, 3 users, load average: 1.31, 1.58, 1.22
      System | innotek GmbH; VirtualBox; v1.2 ()
 Service Tag | 0
    Platform | Linux
     Release | Ubuntu 11.10 (oneiric)
      Kernel | 3.0.0-12-generic
Architecture | CPU = 64-bit, OS = 32-bit
   Threading | NPTL 2.13
     SELinux | No SELinux detected
 Virtualized | VirtualBox
# Processor ##################################################
  Processors | physical = 1, cores = 0, virtual = 1, hyperthreading = no
      Speeds | 1x2240.377
      Models | 1xIntel(R) Core(TM) i3 CPU M 350 @ 2.27GHz
      Caches | 1x6144 KB
# Memory #####################################################
       Total | 1.0M
        Free | 75.2M
        Used | physical = 926.8M, swap allocated = 510.0M, swap used = 44.2M, virtual = 971.0M
     Buffers | 142.6M
      Caches | 481.8M
       Dirty | 116 kB
     UsedRSS | 610.3M
  Swappiness | 60
 DirtyPolicy | 10, 5
 DirtyStatus | 0, 0
  Locator Size Speed Form Factor Type Type Detail
  ========= ======== ================= ============= ============= ===========
# Mounted Filesystems ########################################
  Filesystem Size Used Type Opts Mountpoint
  /dev/sda1 7.4G 52% ext4 rw,errors=remount-ro,commit=0 /
  Downloads 75G 58% vboxsf rw downloads
  none 502M 1% tmpfs rw,noexec,nosuid,nodev,size=5242880 /run/shm
  none 502M 1% tmpfs rw,nosuid,nodev /run/shm
  none 502M 1% debugfs rw /run/shm
  none 502M 1% securityfs rw /run/shm
  none 5.0M 0% tmpfs rw,noexec,nosuid,nodev,size=5242880 /run/lock
  none 5.0M 0% tmpfs rw,nosuid,nodev /run/lock
  none 5.0M 0% debugfs rw /run/lock
  none 5.0M 0% securityfs rw /run/lock
  tmpfs 201M 1% tmpfs rw,noexec,nosuid,size=10%,mode=0755 /run
  udev 495M 1% devtmpfs rw,mode=0755 /dev
# Disk Schedulers And Queue Size #############################
         sda | [cfq] 128
         sr0 | [cfq] 128
# Disk Partioning ############################################
Device Type Start End Size
============ ==== ========== ========== ==================
/dev/sda Disk 8589934592
/dev/sda1 Part 2048 15728639 8052014592
/dev/sda2 Part 15730686 16775167 534774272
/dev/sda5 Part 15730688 16775167 534773248
/dev/sr0 Disk 50776064
# Kernel Inode State #########################################
dentry-state | 17856 10199 45 0 0 0
     file-nr | 5344 0 101176
    inode-nr | 23866 11653
# LVM Volumes ################################################
Unable to collect information
# LVM Volume Groups ##########################################
Unable to collect information
# RAID Controller ############################################
  Controller | No RAID controller detected
# Network Config #############################################
  Controller | Intel Corporation 82540EM Gigabit Ethernet Controller (rev 02)
 FIN Timeout | 60
  Port Range | 61000
# Interface Statistics #######################################
  interface rx_bytes rx_packets rx_errors tx_bytes tx_packets tx_errors
  ========= ========= ========== ========== ========== ========== ==========
  lo 0 0 0 0 0 0
  eth0 100000000 125000 0 4000000 70000 0
# Network Connections ########################################
  Connections from remote IP addresses
  Connections to local IP addresses
  Connections to top 10 local ports
  States of connections
    LISTEN 3
# Top Processes ##############################################
  PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
11165 mysql 20 0 301m 94m 5912 S 9.8 9.5 8:13.34 mysqld
  199 root 20 0 0 0 0 S 3.9 0.0 3:49.19 jbd2/sda1-8
11346 root 20 0 14504 8844 2880 S 3.9 0.9 5:37.23 perl
22824 root 20 0 0 0 0 S 3.9 0.0 0:00.02 kworker/0:1
    1 root 20 0 3320 1628 1264 S 0.0 0.2 0:00.61 init
    2 root 20 0 0 0 0 S 0.0 0.0 0:00.00 kthreadd
    3 root 20 0 0 0 0 S 0.0 0.0 0:01.34 ksoftirqd/0
    5 root 20 0 0 0 0 S 0.0 0.0 0:00.42 kworker/u:0
    6 root RT 0 0 0 0 S 0.0 0.0 0:00.00 migration/0
# Notable Processes ##########################################
  PID OOM COMMAND
    ? ? sshd doesn't appear to be running
# Simplified and fuzzy rounded vmstat (wait please) ##########
  procs ---swap-- -----io---- ---system---- --------cpu--------
   r b si so bi bo ir cs us sy il wa st
   1 0 2 4 20 900 350 700 3 5 69 23
   0 0 25 0 25 1500 800 2500 33 27 12 28
   0 0 0 0 0 1000 500 1000 4 5 59 32
   0 0 0 0 0 1000 450 1000 5 5 60 30
   1 0 0 0 0 900 450 900 4 8 26 62
# The End ####################################################

[Daniel Nichter (daniel-nichter)]

Invalid in 2.0 (I mistakenly thought pt-summary was redesigned in 2.0).

[Daniel Nichter (daniel-nichter)]

I will backport the 2.1 version of shorten() into 2.0 tools so they have a good shoten() too. Since the tools were rewritten from 2.0 to 2.1, the backport is just copying the shorten() func and nothing else.

[Daniel Nichter (daniel-nichter)]

Actually, I'll backport lib/bash/report_formatting.sh and use it entirely in pt-summary and pt-mysql-summary since it's already tested and easy enough to backport.

[Daniel Nichter (daniel-nichter)]

For the record: 2.0.4 doesn't have this bug; it prints 1010.50M.

[Daniel Nichter (daniel-nichter)]

The backport didn't turn out to be so easy because there's little differences like:

# < Checkpoint Age | 0.00
# > Checkpoint Age | 0k

Same value, different representation. Given that this bug isn't actually present in 2.0.4, I'll not fix what isn't broken.

[Daniel Nichter (daniel-nichter)]

Bug 937793 ("pt-mysql-summary : bad size of transaction logs") is related.

[Shahriyar Rzayev (rzayev-sehriyar)]

Percona now uses JIRA for bug reports so this bug report is migrated to: https://jira.percona.com/browse/PT-524