NETGEAR is aware of a growing number of phone and online scams. To learn how to stay safe click here.
Forum Discussion
2TB WD20EARS HCL Advanced Format 4k Sectors TLER LCC WDidle3
EDIT: This thread was written before the Readynas group had recognized and addressed the WD Advanced Format HDD's. An issue that remains to this day is the massive LCC(Load Cycle Count) increases due ...
I would not expect the WD 2.0 TB drives to be able to keep up with the Velociraptors, it is the cargo van vs. a sports car.
We chose the Velociraptors specifically for their low latency, as this NAS runs SQL databases via iSCSI. Surprisingly, the Pro with 6 of these drives via iSCSI is (much) faster than a directly attached Dell PowerVault with 12x 73 GB 10k SCSI drives running off 2 channels on a Perc III RAID controller.
Your system was also slower than the other reports, as it seems the test should take about 6 minutes to run. Since you had 2 GB tests, it should have even been faster (mine was 8 GB test). The 4k sector boundary offset issue would definitely hurt here, but I would only expect about 4 times as long. This problem reminds me of the shoe shining problem slower systems have with fast tape drives.
The weird thing is a contiguous multi-sector write should not incur the penalty nearly as bad, as they should queue up in the drive then get commited. There might be something else at work here causing your troubles. And from what you have mentioned, I would never run a 4k sector drive in a ext4 array unless every drive was a 4k sector drive. The changes in ext4 work well with 4k sectors, but a mix would likely cause nothing but bad performance, or fault intolerance at worst. :shock:
My suggestion would be to replace the two WD drives in your setup (you had 2 of them, and 4 Seagates?). I've had many problems with the Seagate 1.5 TB drives in the past, but hopefully a year+ later they have worked those out.
We chose the Velociraptors specifically for their low latency, as this NAS runs SQL databases via iSCSI. Surprisingly, the Pro with 6 of these drives via iSCSI is (much) faster than a directly attached Dell PowerVault with 12x 73 GB 10k SCSI drives running off 2 channels on a Perc III RAID controller.
Your system was also slower than the other reports, as it seems the test should take about 6 minutes to run. Since you had 2 GB tests, it should have even been faster (mine was 8 GB test). The 4k sector boundary offset issue would definitely hurt here, but I would only expect about 4 times as long. This problem reminds me of the shoe shining problem slower systems have with fast tape drives.
The weird thing is a contiguous multi-sector write should not incur the penalty nearly as bad, as they should queue up in the drive then get commited. There might be something else at work here causing your troubles. And from what you have mentioned, I would never run a 4k sector drive in a ext4 array unless every drive was a 4k sector drive. The changes in ext4 work well with 4k sectors, but a mix would likely cause nothing but bad performance, or fault intolerance at worst. :shock:
My suggestion would be to replace the two WD drives in your setup (you had 2 of them, and 4 Seagates?). I've had many problems with the Seagate 1.5 TB drives in the past, but hopefully a year+ later they have worked those out.
NETGEAR Academy
Boost your skills with the Netgear Academy - Get trained, certified and stay ahead with the latest Netgear technology!
Join Us!
