McFaul

Lol.. does that mean I only get the silver medal? (plus.. in the true internet style.. photos or didn't happen) also sadly very full.....

thats great, thanks!

Oh - an unexpected benefit!!! I now have two controllers.. so Scanner is scanning two disks at once rather than just one as it was before... when you've got 30 disks and 160TB... that makes quite a difference!!! Update: i;ve since realised that i manually reduced the degrees of parallelism while troubleshooting the BSOD, so ive put that back up too!

Well, that was quite an adventure. New board arrived, and looks lovely. Dismantled server, went to move processor from old board to new board, realised the old processor is socket 1150, new board is 1155, (or vice versa, im too lazy to look it up) so it physically wont fit. i order a new processor... while sitting here thinking about it.. i realise i already have a fourth gen i5 in my media pc that i could "borrow" until the new processor arrives... then i realise.... thats a z97 board in my media pc. while its fairly small and there is no room for disks.. i try putting two LSI cards in the media PC, it works perfectly, no issues at all. so i have swapped the motherboard AND processor between my media pc and my server, and voila. Both LSI cards are installed and working perfectly. if i had the physical space i could hang 48 disks off them! now i just need to return the unused "new" motherboard AND the brand new processor that i ordered... and hasnt even been delivered yet!

i'll let you know in 24 hours!

the old one was an Asrock H77 the new one is an Asus Z97 that has several PCI-E 16x slots (neither of them are server boards tho, but if the new one doesnt work i can simply return it for a refund because im an EU person ) https://www.asus.com/uk/Motherboards/Z97AUSB_31/ specifically that one

ha, well hopefully with a new board it will get past bios with the RAID card in the second PCI-E slot (both cards work perfectly in my current board's first PCI-E slot); and i've bought a "better" board

well the HP card arrived Its either not compatible with the LSi card, or its DOA (but it does get warm when its plugged in) but the LSi card doenst recognise it, at all. In the course of troubleshooting - i've also discovered that my motherboard doesnt have a problem with two LSI cards plugged in, it just has a problem with an LSi card being in the second PCI-E slot. with just one LSI card in the second PCI-E slot, and no hard disks, no expanders, no second LSI card, it still wont boot. I'm giving up on the expander now and i've simply ordered a new MB to see if i can get both the LSI cards i now own, to work, since that will give me 48 SATA ports off the two cards.

I ordered the HP - i'll update when it arrives!

I was tempted by the intel one as it uses an LSI chip (and my card is LSi too)

Yep, it works fine with either card, but not both its very frustrating; and im not tempted to change the motherboard looks like i may be keeping one of the cards as a "spare" and have to go down the expander route anyway! any recommendations for expanders? i was looking at the HP 468406-B21

I've even tried totally resetting the BIOS via the clear CMOS jumper.. but still get the same screen with both cards installed

Now i have no BSOD.. it was time to move on to part 2.. which did not go smoothly: http://community.covecube.com/index.php?/topic/1219-ahci-hba-recommendations/

Ok, So now i need some general advice... I have my LSI 9650SE 24 port card happily installed and working. If i install a second one of these cards.. it goes thorough the BIOS for each card fine, detects the drives attached to each drive but the actual motherboard BIOS does not seem happy. it simply goes into a loop of doing this, over and over and over: The motherboard (Asrock H77 PRO4 / MVP) is on the most recent BIOS, and doenst seem to have many other options. With both RAID cards installed you cant even get to the M/B BIOS... it goes DIRECTLY to that screen and when the bar gets to the end, it reboots and starts over!

Yeah, fingers crossed! thanks for your help tho

Well with 2.2.0.628; and the newest firmware on the card... its now been 24 hours of uptime. So its possible whatever drivepool was doing that was managing to crash the previous firmware, doesn't crash the current firmware. I'll post back if it does BSOD, but it seems to have survived past the "danger zone" ! thanks for all your help

Ok, Cheers For now i've put the latest beta on.. since i have the new firmware.. its been up about 12 hours, im expecting a crash in 7 hours unless the new firmware has fixed things if it does, i'll post the Drivepool logs and windows dumps for you and then roll back to an earlier beta and wait another 19 & 1./2 hours!

lol, i wouldn't mind the testing (especially since the issue, so far, only seems to be affecting my card.. but you never know...) .. its the fact that after i install i have to wait almost 24 hours to see if it crashes, lol! so it could be a very slow process.... i'll try the first beta of 2.2 .. if thats stable i may start moving forward in jumps of 10, to at least narrow it down and i'll be sure to keep the drivepool troubleshooting logs on, so hopefully you can see what it was doing right before it died! oh! but i have got the newest firmware on there now.. so maybe the first thing i'll do is try the newest beta and see if maybe the new firmware has fixed it....

Hi, the two you posted in this thread initially this one: http://dl.covecube.com/DrivePoolWindows/beta/download/StableBit.DrivePool_2.2.0.625_x64_BETA.exe which didn't fix the renaming issue; and then this one: http://dl.covecube.com/DrivePoolWindows/beta/download/StableBit.DrivePool_2.2.0.626_x64_BETA.exe Which did fix the renaming issue, but still BSODed

Hi, As i suspected, with 2.1 , its running fine and its been up 24 hours. Its just 2.2 beta which is causing the 3ware driver to BSOD. Which makes it even stranger. Hopefully Alex can see why from the dumps.

No probs! Chris

Hi, I did upload the mini dumps and memory.dmp on the 19th - mcfaul.rar, post 26 in this thread. For the moment it ran for 24 hours with no issues and no scanner or drivepool yesterday until today. Now ive put drivepool 2.1 (12 hours ago) on and am waiting to see if it lasts for 24 hours - which im hoping it does since it ran for more than a week with this version previously. If it does then i'll upgrade to that 2.2 beta on, turn on the drivepool troubleshooting/logging so we can see if that makes it crash. in which case i'll be able to give you the mini dump, the full memory dump, and the drivepoool logs all at once, so hopefully we can see what stranger interaction between the 3ware driver and drivepool is causing BSOD. Chris

So yeah, without drivepool, the server survived the night and has been up for almost 24 hours now. I've updated the firmware and put the release version of drivepool back on (just to see that it isnt drivepool per se, but the latest beta, which was when the crashing started) So lets see if it can survive 24 hours with just ordinary drivepool on.

I did see a bunch of driver related errors in event log; ive uploaded the event log, not that it tells you a huge deal (mcfaul.evtx); there's no hint of whats triggering it, just that the driver crashes! The card defintiely isnt overheating - its 35C and has a fan blowing directly on it. However i have realised the card actually has a fairly out of date firmware! (D'Oh) im sure i checked it was the latest but i must have misread since its got 4.10.00.021 and the latest ends 27... I'm not going to restart it now, but i'll update it tomorrow and see if that affects it. Right, bedtime, lets see what i find in the morning!

Yeah it is super odd. At the moment its been up exactly 12 hours, so i'll hold off doing anything until tomorrow when it will have been up for 24 hours (hopefully!). I do use it for video conversion sometimes, which maxes out all 4 cores at 100%, so that'll make a handy CPU stress test. The system drive is a samsung SSD and shows as being healthy, but that will be easy to test. I cant see anything in task scheduler which looks out of place, or that triggers either anywhere near the time of the crashes, or after about 19 hours and 20 minutes of uptime! Did Alex get a chance to look at the dumps? (Since i have no idea what im doing with those). Anyway, almost bedtime here, i'll post again in the morning if its made it through the night, or not...