Thanks for the fast answer!
The temperature thing is a good hint, I will have an eye on that (although my modules are good vented with some space on the side).
From my side I wasn't able to pinpoint it down to a specific source of error, sometimes it happens more often and sometimes less.
According to my own experience and a search on the web possible sources could be power condition/filtering, IC2 and as you temperature.
What i discovered is that a soft reset (reboot on the module) works much less than a reboot done on the whole rack with waiting a few seconds before power it on again.
While exchanging the RAM would theoretically be possible, it doesn't make sense to me do it as long as we dont know the real reason for the error.
Would be interesting to hear what OS is doing on repair, just changing the IC or do some additional measures on the electrical side aeg adding filtering caps?
Did he go deeper on the issues and a possible solution when you've been in contact with him?
Im only using them in single mode, so dual mode is not an option for me...
Thanks and cheers!