P70 - Raid 1 PCIE-NVME (512GB) -- horrid and random IO lag

[jackdpeterson]

I've been fighting this intermittently since I got this laptop. At times, the laptop runs perfectly fine. Other times the system hangs until it's good and ready to respond.

When I open up the performance monitor I notice that disk Activity meter is at 100% with varying levels of Read / Write IO (typically in the tens of MBps), though it could be lower.

Obviously this is nowhere near what the disks are specified for. Does anyone have any suggestions on how to validate what's wrong as far as diagnostic tools go?

Chkdisk comes back clean w/o any bad sectors (though I think that could be masked by the controller and the magic of SSDs re-allocating storage to non-bad sectors).

I recently also did a clean reinstall of windows and the problems persist. At first I thought this was maybe a bug with the Intel RST drivers ... so I did the complete clean reinstall and that didn't help. Very frustrated at this point as the tooling available doesn't have diagnostics for hardware. Lenovo's online diagnostic tool indicated 'Device not support'. And there are no BIOS-level diagnostics that I've observed as being available to run.

System specs: P70 20ER

Xeon 1575M

64GB ECC

RAID 1 option w/ two pcie-nvme SSDs (the Samsung 950 or the 951's I believe -- whatever Lenovo installs).

4k screen w/ the M5000M GPU

Qualcomm chip installed.

Also, firmware on the SSDs is patched per the available lenovo update site.

Latest win 10 x64 build as of the date of this writing w/ all updates installed.

At this point, any suggestions would really be appreciated as I'm at a loss as to what to do at this point. I'd really like to get to the bottom of what's going on.

[richk]

I like crystaldiskinfo (freeware) to check drive health as a first step.  I then keep task manager running to see what is using all the disk reads or writes.

[jackdpeterson]

Hmm, gave CrystalDiskInfo a shot -- enabled the advanced detetion as well and it can't detect the 'disks' (though I suppose it would be a controller). guess I'll ping the official tech support side of things to see what they propose and will keep this thread updated as to progress.

[ColonelONeill]

Did you also check to see if your Intel Rapid Storage drivers and RAID firmware are up to date?

[richk]

I admit I have never tried CrystalDiskInfo on an internal RAID.  The only RAID I am using is on a NAS device hooked to my desktop.  It cannot be seen either.

On the function menu tab/advanced features. there are Intel RAID options, but I have not used them.

[jackdpeterson]

Yes - the RST Driver and storage firmware are up to date. I just installed (re-installed as I had this prior to doing a clean-install of Windows). No updates were needed as far as the firmware goes as I've been very active in watching the drivers rolling out for the P70. RST is re-running verification and I don't anticipate it to find any problems (as it hasn't in the past).

What's interesting is that despite having RST installed I can't get any S.M.A.R.T. information out of the disks individually. Even w/ the addition of the RST driver + application installed and drive firmware at latest, Crystal Disk Info cannot detect the controller / drives.

Is there a more advanced / developer friendly set of debugging tools that one can use with the Intel RST controller to gain information about the underlying drives?

On a related note, I did discover a solution to part of my problem (just while using VMWare Workstation) which is to ensure that all VM memory is stuffed in RAM rather than relying on disk IO ... so that helps to reduce some of the issues realted to IO clogging. 

General windows still hangs intermittently -- though I'll provide an update on that now that I've reinstalled the RST driver and verified firmware is current and consistent between both drives.

[ColonelONeill]

Here's an oddly written Intel KB article that I'm not really sure explains anything:
https://www.intel.com/content/www/us/en/support/technologies/000006374.html

Does the Intel Rapid Storage Manager list drive information correctly?

Also, what are the SSD temperatures under load? Maybe they're throttling:
https://www.pugetsystems.com/labs/articles/Samsung-950-Pro-M-2-Throttling-Analysis-776/#Worstcase%28...