Re: Tainted kernel possible caused by postgresql

On Tue, Sep 29, 2015 at 8:19 AM, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
> KK <krishnakumar(dot)143(at)gmail(dot)com> writes:
>> Sep 28 23:31:07 rac-101 kernel: [223791.090456] CPU: 27 PID: 30123 Comm:
>> postgres Tainted: G W 3.14.53-20150923-0842-cc65b30 #1
>
> There was a previous discussion of this message which concluded that it
> must be a kernel bug:
> http://www.postgresql.org/message-id/flat/CALXbVnehCf4XX1umL=ZJ+RJ5OO_DQagNiVubt48h0B6oOL2swQ(at)mail(dot)gmail(dot)com
>
> I'd suggest reporting it to your kernel vendor and see what they say.

Also, when reporting kernel bugs:
*) Whether or not the system is virtualized (and if so, by which VM
system) is extremely relevant
*) Please report O/S vendor and version
*) Consider listing loaded kernel modules. Exotic storage and network
drivers are especially interesting (say, fiber channel HBA or TCP
offload).
*) Research similar reports both internally within the archives and in
general. If anything is found supply it as context with the report
*) It's helpful to rule out hardware problems. If it's possible to do
so, run some hardware diagnostics

check out this thread and see if it applies (via google search,
warn_slowpath_common tainted)
https://www.centos.org/forums/viewtopic.php?t=3642

merlin