Eh, I don't think so. I've been running SMP for years. This particular
machine is a very idle machine with only simple small stuff going on
with developers. I have another SMP machine in the same configuration
doing millions of rows with quite a load on it. If it were VM code
fault, I would have expected it to occur over time and be corrupted
data, not transferred data.
BTW, it wasn't secretly gutted, if you read LKML you'd have known about
the plans well in advance. :)
David
p.s. I haven't had filesystem corruption for 3 years except for a real
serious power surge/brown two weeks ago which blew out a UPS, power
supply, and one motherboard. All systems remained intact except for the
harddrive under the powersupply. My fs corruption 3 years ago was due to
repeated battery failures on a laptop.
Michael Loftis wrote:
> You might want to try chekcign with the Linux guys. They have a bad
> habit of getting filesystem corrupting bugs in their SMP code. And
> being as the VM wa srecently gutted in the 'stable' kernel it almost
> sounds like something wrong with the Linux VM.
>
> David Ford wrote:
>
>> In my haste to get an answer, I ommitted certain needful facts.
>>
>> I'm running on an SMP Linux 2.4.19p6, w/ pgsql 7.2. This
>> installation was running fine since March 16th.
>>
>> David
>>
>>
>>
>> ---------------------------(end of broadcast)---------------------------
>> TIP 2: you can get off all lists at once with the unregister command
>> (send "unregister YourEmailAddressHere" to majordomo@postgresql.org)
>
>
>