dear all,
can anyone help me.
I have a backup destination storage server with the following specs:
HP DL180 G6
P410 smart array controller with 1GB cache, 6.64 firmware, 75% read 25% write cache settings with NO battery however it is set to still enable the cache when no battery is present
3 x Disk Arrays are configured, array 1 is for the operating system and has 2 x 4TB disks in RAID 1
the other 2 x arrays each have 4 x 8TB disks with a RAID 10 setup
Windows 2012 R2
48GB RAM
2 x Intel E5620 cpus
the 2 x non OS disk arrays which are used to store the backup files are encrypted with Bitlocker and have dedup enabled to run during the day when no backups are running. Each disk array is setup as a 14.5TB NTFS disk. Now I realise the recommended max size for a windows 2012 r2 dedup disk is 10TB and mine is 14.5TB but I don't think this is related to the issue, but I suppose it is an orangey-red flag.
So basically when I run a disk speed test program checking for the writes and reads I normally get around 300MB both ways which I am happy about considering they are encrypted etc I have also run this during backups and I still get around 300MB.
So the problem I have is after a full night's worth of backups and often when backups are still running, occasionally I get very very poor disk speed checks on writes and reads sometimes like 5MB! This will cause the backups to be delayed and in some cases fail which is not good. Also worth mentioning when I run the disk speed program I also get poor results on the OS disk array which is not encrypted or deduped, so thats why I'm thinking it related to the disk controller.
Is it possible the HP Smart Array cache is running out of memory or that there is a memory leak somewhere - possibly relating to the firmware version?
I also get this problem occasionally on another almost identical backup storage server, so I don't think it is a hardware fault.
Any help is much appreciated.
Mike
