dumpsys -s reports huge space requirements.

[Christof Schoeman]

Hi

I need to run a dumpsys for a case being investigated, but this is what dumpsys says it needs:

#dumpsys -s
Approximate full dump size = 910163968 disk blocks,
if compressed, expect about 455081984 disk blocks.
Approximate partial dump size = 6884784 disk blocks,
if compressed, expect about 3442392 disk blocks.

Do I really need about 217GB for a compressed full dump?

If so, how is that possible if the system only has 32GB of memory and hardly any swap space is being used?

[Ivan Ferreira]

If i'm not wrong, for a full dump you only need 444 MB.

Why do you say 217 GB?

Remember that is reported in 512-byte blocks.

[Michael Schulte zur Sur]

Hi,

there is something wrong with the figures.
444mb would be 444*1024*2 blocks which is 909312 blocks. So the numbers are off by a factor of 1000.

greetings,

Michael

[Christof Schoeman]

How I got to 217GB...

"if compressed, expect about 455081984 disk blocks"

455081984 blocks / 2 = 227540992 KB
227540992 KB / 1024 = 222208 MB
222208 MB / 1024 = 217 GB

So, 217GB seems excessive and 222MB (if out by a factor of 1000) seems too little for a 32GB memory system (being 90% utilized).

I'm hesitant to do a dumpsys, just in case the figures are correct.

[Ivan Ferreira]

Yep, I was wrong, I divided by 2048 but my reading was wrong.

How much swap space do you have configured? Are you using swap space? swapon -s

How is the memory allocation configured? see vm_swap_eager in /etc/sysconfigtab

Also, remember that the exact value cannot be calculed with dumpsys. See man dumpsys.

[Christof Schoeman]

Total swap space = 66.43GB
Used = 1%

vm_swap_eager = 0

I'm aware that dumpsys cannot give an accurate figure, but this is way off.

Think I'll just make sure there is anough disk space to take the total physical memory, plus the space used in swap.

[Venkatesh BL]

What is the system configuration?. Is it a NUMA system?

[Christof Schoeman]

System is a 16 CPU GS1280.

[Venkatesh BL]

Is 32 GB the total RAM present on the system. Can you post the output of 'vmstat -P'?

[Christof Schoeman]

vmstat -P

Total Physical Memory = 32768.00 M
= 4194304 pages

Physical Memory Clusters:

start_pfn end_pfn type size_pages / size_bytes
0 1984 pal 1984 / 15.50M
1984 262144 os 260160 / 2032.50M
2097152 2097212 pal 60 / 480.00k
2097212 2359296 os 262084 / 2047.53M
4194304 4194364 pal 60 / 480.00k
4194364 4456448 os 262084 / 2047.53M
6291456 6291516 pal 60 / 480.00k
6291516 6553600 os 262084 / 2047.53M
16777216 16777276 pal 60 / 480.00k
16777276 17039360 os 262084 / 2047.53M
18874368 18874428 pal 60 / 480.00k
18874428 19136512 os 262084 / 2047.53M
20971520 20971580 pal 60 / 480.00k
20971580 21233664 os 262084 / 2047.53M
23068672 23068732 pal 60 / 480.00k
23068732 23330816 os 262084 / 2047.53M
33554432 33554492 pal 60 / 480.00k
33554492 33816576 os 262084 / 2047.53M
35651584 35651644 pal 60 / 480.00k
35651644 35913728 os 262084 / 2047.53M
37748736 37748796 pal 60 / 480.00k
37748796 38010880 os 262084 / 2047.53M
39845888 39845948 pal 60 / 480.00k
39845948 40108032 os 262084 / 2047.53M
50331648 50331708 pal 60 / 480.00k
50331708 50593792 os 262084 / 2047.53M
52428800 52428860 pal 60 / 480.00k
52428860 52690944 os 262084 / 2047.53M
54525952 54526012 pal 60 / 480.00k
54526012 54788096 os 262084 / 2047.53M
56623104 56623164 pal 60 / 480.00k
56623164 56885248 os 262084 / 2047.53M

Physical Memory Use:

start_pfn end_pfn type size_pages / size_bytes
1984 2057 scavenge 73 / 584.00k
2057 2983 text 926 / 7.23M
2983 3072 scavenge 89 / 712.00k
3072 3253 data 181 / 1.41M
3253 3482 bss 229 / 1.79M
3482 3700 kdebug 218 / 1.70M
3700 3707 cfgmgmt 7 / 56.00k
3707 3709 locks 2 / 16.00k
3709 3726 pmap 17 / 136.00k
3727 12263 unixtable 8536 / 66.69M
12263 12647 logs 384 / 3.00M
12647 20363 vmtables 7716 / 60.28M
20363 262143 managed 241780 / 1888.91M
*
2097152 2097213 unixtable 61 / 488.00k
2097213 2097216 pmap 3 / 24.00k
2097216 2097673 scavenge 457 / 3.57M
2097673 2098599 text 926 / 7.23M
2098599 2104570 vmtables 5971 / 46.65M
2104570 2359295 managed 254725 / 1990.04M
*
4194304 4194365 unixtable 61 / 488.00k
4194365 4194368 pmap 3 / 24.00k
4194368 4194825 scavenge 457 / 3.57M
4194825 4195751 text 926 / 7.23M
4195751 4201722 vmtables 5971 / 46.65M
4201722 4456447 managed 254725 / 1990.04M
*
6291456 6291517 unixtable 61 / 488.00k
6291517 6291520 pmap 3 / 24.00k
6291520 6291977 scavenge 457 / 3.57M
6291977 6292903 text 926 / 7.23M
6292903 6298874 vmtables 5971 / 46.65M
6298874 6553599 managed 254725 / 1990.04M
*
16777216 16777277 unixtable 61 / 488.00k
16777277 16777280 pmap 3 / 24.00k
16777280 16777737 scavenge 457 / 3.57M
16777737 16778663 text 926 / 7.23M
16778663 16784634 vmtables 5971 / 46.65M
16784634 17039359 managed 254725 / 1990.04M
*
18874368 18874429 unixtable 61 / 488.00k
18874429 18874432 pmap 3 / 24.00k
18874432 18874889 scavenge 457 / 3.57M
18874889 18875815 text 926 / 7.23M
18875815 18881786 vmtables 5971 / 46.65M
18881786 19136511 managed 254725 / 1990.04M
*
20971520 20971581 unixtable 61 / 488.00k
20971581 20971584 pmap 3 / 24.00k
20971584 20972041 scavenge 457 / 3.57M
20972041 20972967 text 926 / 7.23M
20972967 20978938 vmtables 5971 / 46.65M
20978938 21233663 managed 254725 / 1990.04M
*
23068672 23068733 unixtable 61 / 488.00k
23068733 23068736 pmap 3 / 24.00k
23068736 23069193 scavenge 457 / 3.57M
23069193 23070119 text 926 / 7.23M
23070119 23076090 vmtables 5971 / 46.65M
23076090 23330815 managed 254725 / 1990.04M
*
33554432 33554493 unixtable 61 / 488.00k
33554493 33554496 pmap 3 / 24.00k
33554496 33554953 scavenge 457 / 3.57M
33554953 33555879 text 926 / 7.23M
33555879 33561850 vmtables 5971 / 46.65M
33561850 33816575 managed 254725 / 1990.04M
*
35651584 35651645 unixtable 61 / 488.00k
35651645 35651648 pmap 3 / 24.00k
35651648 35652105 scavenge 457 / 3.57M
35652105 35653031 text 926 / 7.23M
35653031 35659002 vmtables 5971 / 46.65M
35659002 35913727 managed 254725 / 1990.04M
*
37748736 37748797 unixtable 61 / 488.00k
37748797 37748800 pmap 3 / 24.00k
37748800 37749257 scavenge 457 / 3.57M
37749257 37750183 text 926 / 7.23M
37750183 37756154 vmtables 5971 / 46.65M
37756154 38010879 managed 254725 / 1990.04M
*
39845888 39845949 unixtable 61 / 488.00k
39845949 39845952 pmap 3 / 24.00k
39845952 39846409 scavenge 457 / 3.57M
39846409 39847335 text 926 / 7.23M
39847335 39853306 vmtables 5971 / 46.65M
39853306 40108031 managed 254725 / 1990.04M
*
50331648 50331709 unixtable 61 / 488.00k
50331709 50331712 pmap 3 / 24.00k
50331712 50332169 scavenge 457 / 3.57M
50332169 50333095 text 926 / 7.23M
50333095 50339066 vmtables 5971 / 46.65M
50339066 50593791 managed 254725 / 1990.04M
*
52428800 52428861 unixtable 61 / 488.00k
52428861 52428864 pmap 3 / 24.00k
52428864 52429321 scavenge 457 / 3.57M
52429321 52430247 text 926 / 7.23M
52430247 52436218 vmtables 5971 / 46.65M
52436218 52690943 managed 254725 / 1990.04M
*
54525952 54526013 unixtable 61 / 488.00k
54526013 54526016 pmap 3 / 24.00k
54526016 54526473 scavenge 457 / 3.57M
54526473 54527399 text 926 / 7.23M
54527399 54533370 vmtables 5971 / 46.65M
54533370 54788095 managed 254725 / 1990.04M
*
56623104 56623165 unixtable 61 / 488.00k
56623165 56623168 pmap 3 / 24.00k
56623168 56623625 scavenge 457 / 3.57M
56623625 56624551 text 926 / 7.23M
56624551 56630522 vmtables 5971 / 46.65M
56630522 56885247 managed 254725 / 1990.04M
============================
Total Physical Memory Use: 4192303 / 32752.37M

Managed Pages Break Down:

free pages = 1208295
active pages = 1352113
inactive pages = 589018
wired pages = 454069
ubc pages = 466162
==================
Total = 4069657

WIRED Pages Break Down:

vm wired pages = 170334
ubc wired pages = 39
meta data pages = 125689
malloc pages = 111287
contig pages = 2292
user ptepages = 2865
kernel ptepages = 9
free ptepages = 11
==================
Total = 412526