My Oracle Support Banner

Oracle Linux 7: Kernel Memory Leak for Veritas File System (Doc ID 2168986.1)

Last updated on MARCH 27, 2019

Applies to:

Linux OS - Version Oracle Linux 7.0 and later
Oracle Cloud Infrastructure - Version N/A and later
Linux x86-64

Symptoms

Oracle Database becomes unresponsive, and below symptoms are noticed on OL7 with Veritas file system.

1. Out of Memory in OS logs
2. High usage of slab
3. Huge number of objects for kmalloc-1024

<hostname> kernel: Out of memory: Kill process 40247 (oracle) score 0 or sacrifice child
<hostname> kernel: Killed process 40247 (oracle) total-vm:157584396kB, anon-rss:0kB, file-rss:1648kB
<hostname> kernel: oracle invoked oom-killer: gfp_mask=0x42d0, order=3, oom_score_adj=0
<hostname> kernel: oracle cpuset=/ mems_allowed=0-3
<hostname> kernel: CPU: 6 PID: 112912 Comm: oracle Tainted: P OE ------------ 3.10.0-327.10.1.el7.x86_64 #1
<hostname> kernel: Hardware name: Dell Inc. PowerEdge R920/0Y4CNC, BIOS 1.3.2 09/30/2014
<hostname> kernel: ffff88240bdb0000 000000000a4eeef6 ffff880cf48a7910 ffffffff816352a4
<hostname> kernel: ffff880cf48a79a0 ffffffff8163023f ffff882e3a1e7fa0 ffff882e3a1e7fb8
<hostname> kernel: ffffffff00000202 fafeebff00000000 0000000000000002 ffffffff81128903
<hostname> kernel: Call Trace:
<hostname> kernel: [<ffffffff816352a4>] dump_stack+0x19/0x1b
<hostname> kernel: [<ffffffff8163023f>] dump_header+0x8e/0x214
<hostname> kernel: [<ffffffff81128903>] ? delayacct_end+0x83/0xb0
<hostname> kernel: [<ffffffff8116ce4e>] oom_kill_process+0x24e/0x3b0
<hostname> kernel: [<ffffffff81088d8e>] ? has_capability_noaudit+0x1e/0x30
<hostname> kernel: [<ffffffff8116d676>] out_of_memory+0x4b6/0x4f0
<hostname> kernel: [<ffffffff81173855>] __alloc_pages_nodemask+0xa95/0xb90
<hostname> kernel: [<ffffffff811b4429>] alloc_pages_current+0xa9/0x170
<hostname> kernel: [<ffffffff81514bc0>] sk_page_frag_refill+0x70/0x160
<hostname> kernel: [<ffffffff81576ca3>] tcp_sendmsg+0x263/0xc20
<hostname> kernel: [<ffffffff815a1074>] inet_sendmsg+0x64/0xb0
<hostname> kernel: [<ffffffff812889f3>] ? selinux_socket_sendmsg+0x23/0x30
<hostname> kernel: [<ffffffff8150ff37>] sock_aio_write+0x157/0x180
<hostname> kernel: [<ffffffff811ddd7d>] do_sync_write+0x8d/0xd0
<hostname> kernel: [<ffffffff811de695>] vfs_write+0x1b5/0x1e0
<hostname> kernel: [<ffffffff811df03f>] SyS_write+0x7f/0xe0
<hostname> kernel: [<ffffffff81645a49>] system_call_fastpath+0x16/0x1b

slab :

zzz ***
MemTotal: 528142476 kB
MemFree: 789636 kB
MemAvailable: 995184 kB
Slab: 234003392 kB

Huge number of objects is used for kmalloc-1024  :

zzz ***
slabinfo - version: 2.1
# name <active_objs> <num_objs> : tunables : slabdata <active_slabs> <num_slabs>
kmalloc-1024 231175235 231175452 1024 32 8 : tunables 0 0 0 : slabdata 7601756 7601756 0
kmalloc-512 1569361 1623456 512 32 4 : tunables 0 0 0 : slabdata 50739 50739 0

Changes

No changes

Cause

To view full details, sign in with your My Oracle Support account.

Don't have a My Oracle Support account? Click to get started!


In this Document
Symptoms
Changes
Cause
Solution
References


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