OVM: Oracle VM DomU HVM (HVM+PV) image restart - p2m_pod_demand_populate: Out of populate-on-demand memory! (Doc ID 1444613.1)

Last updated on NOVEMBER 07, 2023

Applies to:

Symptoms

A Oracle VM HVM (HVM+PV) image (Dom) is crashing or restarting.

DomU's restart time can be seen from /var/log/messages in guest OS:

Mar 20 11:13:13 hostdomu xinetd[3480]: START: bgssd pid=6079 from=xxx.xxx.xxx.xxx
Mar 20 11:13:13 hostdomu xinetd[3480]: EXIT: bgssd status=0 pid=6079 duration=0(sec)
Mar 20 13:28:38 hostdomu syslogd 1.4.1: restart.
Mar 20 13:28:38 hostdomu kernel: klogd 1.4.1, log source = /proc/kmsg started.
Mar 20 13:28:38 hostdomu kernel: Linux version 2.6.18-238.9.1.0.1.el5 (mockbuild@ca-build9.xx.yy.com) (gcc version 4.1.2 20080704 (Red Hat 4.1.2-50)) #1 SMP Tue Apr 12 19:58:35 EDT 2011
Mar 20 13:28:38 hostdomu kernel: Command line: ro root=/dev/rootvg/rootvol ide0=noprobe ide1=noprobe
Mar 20 13:28:38 hostdomu kernel: BIOS-provided physical RAM map:
Mar 20 13:28:38 hostdomu kernel: BIOS-e820: 0000000000010000 - 000000000009fc00 (usable)

On the VM Server Dom0 in the /var/log/xen/xend.log:

[2012-03-20 13:27:59 5876] WARNING (XendDomainInfo:2132) Domain has crashed: name=0004fb00000600002f0b4194eef3b3c8 id=41.   <--DomU name in vm.cfg, note id# 
[2012-03-20 13:27:59 5876] DEBUG (XendDomainInfo:3118) XendDomainInfo.destroy: domid=41
[2012-03-20 13:28:01 5876] DEBUG (XendDomainInfo:2474) Destroying device model
[2012-03-20 13:28:01 5876] INFO (image:612) 0004fb00000600002f0b4194eef3b3c8 device model terminated
[2012-03-20 13:28:01 5876] DEBUG (XendDomainInfo:2481) Releasing devices
[2012-03-20 13:28:01 5876] DEBUG (XendDomainInfo:2487) Removing vif/0
[2012-03-20 13:28:01 5876] DEBUG (XendDomainInfo:1297) XendDomainInfo.destroyDevice: deviceClass = vif, device = vif/0
[2012-03-20 13:28:01 5876] DEBUG (XendDomainInfo:2487) Removing vif/1
[2012-03-20 13:28:01 5876] DEBUG (XendDomainInfo:1297) XendDomainInfo.destroyDevice: deviceClass = vif, device = vif/1

On the VM Server Dom0 in #xm dmesg log we see same domain id 41 :

(XEN) p2m_pod_demand_populate: Out of populate-on-demand memory! tot_pages 2018296 pod_entries 4034560
(XEN) domain_crash called from p2m.c:1127
(XEN) Domain 41 (vcpu#7) crashed on cpu#0:
(XEN) ----[ Xen-4.0.2-OVM x86_64 debug=n Not tainted ]----
(XEN) CPU: 0
(XEN) RIP: 0033:[<00002aaaabd5e00b>]
(XEN) RFLAGS: 0000000000010297 CONTEXT: hvm guest
(XEN) rax: 00002aaabb751005 rbx: 00000000fbf25f20 rcx: 0000000000000000
(XEN) rdx: 00000000fbf29fd8 rsi: 00000000fbf23e30 rdi: 0000000000000002
(XEN) rbp: 00000000fbf23df8 rsp: 00000000430458d0 r8: 0000000000000002
(XEN) r9: 00000000fbf23e30 r10: 00000000fbf29ff0 r11: 0000000000000005
(XEN) r12: 0000000000000000 r13: 00000000fbf23e30 r14: 00000000fbf25f00
(XEN) r15: 00002aaabb751000 cr0: 0000000080050033 cr4: 00000000000006e0
(XEN) cr3: 0000000561dc1000 cr2: 00002aaaaacff000
(XEN) ds: 0000 es: 0000 fs: 0063 gs: 0000 ss: 002b cs: 0033
(XEN) grant_table.c:1998:d0 destination frame ffffffffffffffff invalid.
(XEN) grant_table.c:1998:d0 destination frame ffffffffffffffff invalid.
(XEN) grant_table.c:1998:d0 destination frame ffffffffffffffff invalid.

In VM Server 3.4 and higher the messages look like:

(XEN) p2m_pod_demand_populate: Dom30 out of PoD memory! (tot=4195319 ents=4194304 dom0)
(XEN) domain_crash called from ffff82d08030d9cb
(XEN) p2m_pod_demand_populate: Dom30 out of PoD memory! (tot=4195319 ents=4194304 dom0)
(XEN) domain_crash called from ffff82d08030d9cb
(XEN) p2m_pod_demand_populate: Dom30 out of PoD memory! (tot=4195319 ents=4194304 dom0)
(XEN) domain_crash called from ffff82d08030d9cb

Cause

My Oracle Support provides customers with access to over a million knowledge articles and a vibrant support community of peers and Oracle experts.