We have an Intel Server (as mentioned in title) with the following config:
16 GB RAM
1 CPU Intel Xeon E5-2630
4 x Hitachi HUC109090CSS600 in RAID 6, with two partitions: C (300GB) and D (1.34TB)
APC Smart-UPS 2200
Windows Server 2012 Standard
In Intel Raid Web Console, the setup is as follows:
- Disk Cache Policy: Unchanged
- Read Policy: Always Read Ahead
- IO Policy: Direct IO
- Write Policy: Write Back (with default policy - Always Write Back, as we did not purchase a BBU yet)
For a week since we installed this server, everything went smoothly. Yesterday evening the Event Viewer logged the following events (in order of appearance):
- 12 x The speed of processor 0 in group 0 is being limited by system firmware. The processor has been in this reduced performance state for 71 seconds since the last report. = 1 warning for each logical CPU
- 3 x While transmitting or receiving data, the server encountered a network error. Occasional errors are expected, but large amounts of these indicate a possible error in your network configuration. The error status code is contained within the returned data (formatted as Words) and may point you towards the problem. - which I assume is from a short power failure that happened around that time - we have some switches in other locations, not backed up by UPSes.
Since then, the disk access became slower than a normal SATA disk, and all activity is just sluggish. Even launching the Event Viewer takes a lot of time. I've seen this performance before switching from Write Thru to Write Back.
I've rarely seen the CPU going above 20% and the memory has a constant rate of 36% usage.
Copying a 4GB iso from D to C:
Also, two benchmarks made, one last week and one now. You can guess wich is which
Allow me to answer myself.
I waited until everybody went home, restarted the server to check RAID console bios. And the weird thing is, although RAID Web Console was reporting Write Back w/ Always Write Back, in RAID Bios it was Write Thru.
So, I assume it's a RAID Web Console bug.