Can you capture "show memory main-heapsh” before and after?
Will show you exact number of pages mapped.
I.e.
vpp# show memory main-heap
Thread 0 vpp_main
virtual memory start 0x7f13c03ae000, size 1048640k, 262160 pages, page size 4k
numa 1: 45220 pages, 180880k
not mapped: 216940 pages, 867760k
total: 1.00G, used: 43.19M, free: 980.88M, trimmable: 978.81M
This means that 45220 pages (180MB) out of 262160 (1G of 4K pages) is mapped.
New page is dynamically allocated by kernel 1st time we write a byte into its
memory address.
In VPP page is never freed, even if we free all allocations which hold that
page, kernel still thinks it is used by us
and VPP memory allocator will reuse it.
Effect of this is that if you allocate 800Mb out of heap, touch at least one
byte at each page, and then you free that allocation then
freed memory will be available for other allocations, but from kernel
perspective 800MB of pages will stay in use for a VPP lifetime ...
—
Damjan
> On 1 Apr 2020, at 14:58, Satya Murthy <satyamurthy1...@gmail.com> wrote:
>
> Few questions on VPP memory usage.
>
> 1. Using pmap -p <pid of vpp> i am collecting the total memory usage of vpp
> process at the beginning of my test. ( it is X KB )
> 2. I ran test for few hours which will obviously have lot of pool_get/pool_put
> 3. Collected the same same pmap output and the memory usage grown to X+Y
> 4. Cleared all my sessions which ideally should have cleared all my
> allocations.
> 5. I still see the memory usage being stuck at X+Y
>
> I see that only pool_free is doing unmap of the memory, whereas pool_put is
> not doing so.
> Could this be the reason why I am seeing memory not coming down.
>
> Thanks & Regards,
> Satish
>
>
> --
> Thanks & Regards,
> Murthy
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#15969): https://lists.fd.io/g/vpp-dev/message/15969
Mute This Topic: https://lists.fd.io/mt/72699704/21656
Group Owner: vpp-dev+ow...@lists.fd.io
Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub [arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-
