Thank you for the knowledge share! I'm new to the channel, but have learned a lot thus far. Now, I have a complete VNX array that I have been looking to sell, but since seeing your channel, I have decided that I may want to repurpose a shelf or two to support a TrueNAS install. I do not have a server yet, but will continue researching a best fix - along with the recommendations you have made. Is there a particular RAID card I should target? I'd like to run no more than 2 shelves from one server and would like to keep it low power as much as reasonably possible. Again, thanks for the insight and inspiration!
Do you have a video for testing hard drives? Also, would you happen to know how to test EMC, NetApp, and other hard drives? I have used HP and Dell servers to test the drives using their own software, but I could not figure out EMC and other drives, especially those fiber channel drives.
Yes I do and yes I have... The easiest way to do this is with a USB stick with Linux on it and use my videos to reformat and testing your disks as needed. Also don't over look your SAS controller diag tools that can diag disks on the fly plus reformat and fast int. your drives. Your RAID and HBA controllers run diags on your disks at level one and level two and will fail a drive if they fail those post BIOS diag checks. The only thing that the controllers will not do is the level 3 HDD RAM cache test. That is the cache on the drive and that will require a bandwidth testing tool from Crystal reports HDD tool set but that is all you need. Check out my video on reformating a disk using Linux from sector 520 (EMC and NetApp formats) over to PC satandard for PC and x86 servers as 512 sector formatting. Also check out my video on how to do this type of reformating on scale with many disks at once to speed things up. Hope this helps! And enjoy your fun!
Well first your HBA or RAID Controller run a series of DIAG tests for pass of fail and that is key and fastest to do. next you can use the basic DMS (Disk Management System) tools of windows or Linux to do the basic testing and formatting. Lastly you can use the Crystal disk reporting tools to some level. You see these disks (Most of them) are enterprise level disks and run in a different format. It is not like a PC hard drive. Those PC based disk tools will work but you have to place all of your enterprise drives as RAID 0 or stand alone mode and that is a lot of work. SO recommend the tools that I use. Start with your HBA or RAID controller first to see if you have an issue or not and these controllers are really good at doing this testing. Then work from there. Hope this helps some... Keep enjoying this stuff as you learn more
On PC I've used hdsentinal's surface restore function (write and check AA, 55, AA, 00 to every byte on the disk). Each sector is measured for response time after resurfacing so identify any weak sectors after surface restore. Then it reports the disks overall health. So, not relying solely on S.M.A.R.T. quick or long scan to tell me the overall health of the drive. The process can take a looooong time, like days. I was hoping you had discovered an automated way do something similar with a disk array setup and hot swap bays. Like plug in one or more drives. System automatically queues it up to be scanned/restored. As each surface restore is completed it prints out a health report label on an attached label printer. Then each morning I look at the labels printed out for the drives that have completed, remove them, apply the label, put in another drive, etc. No need to type in any commands during any of the drive scanning.
Whats your experience with CX2 ssd or the adata su650? The prices of these ssd have dropped to a level that i can consider them to replace my 900gb sas.
I have used them both. I use the CX2 1TB and 2 TB with my disk arrays and the su650 as well but mostly for system boot drives for you can remove them from the 2.5 casing and the SATA SSD board is really small. Fun to work with them that way but my main say is the CX2 when I buy in bulk SSD drives. Hope that helps some.
hey I use a food grade tub with silica gel packets for my excess drives so moisture doesn't kill them and if a HDD in my NAS dies I just go in and look for a stack with the same storage size if none then the array may gain more space.
Hit the nail on the head with the organisation (even if it's not fun). If your gear is ready to go, tested and tidy - you will lab more.
No worries, good pearls of wisdom from experience is always nice to hear and those who understands will appreciate them for sure! :)
Thank you for the knowledge share! I'm new to the channel, but have learned a lot thus far. Now, I have a complete VNX array that I have been looking to sell, but since seeing your channel, I have decided that I may want to repurpose a shelf or two to support a TrueNAS install.
I do not have a server yet, but will continue researching a best fix - along with the recommendations you have made.
Is there a particular RAID card I should target? I'd like to run no more than 2 shelves from one server and would like to keep it low power as much as reasonably possible.
Again, thanks for the insight and inspiration!
I'd love to see all of the other aspects of a home lab, that's what's missing from other videos
Well shucks! Thank you very much and keep having fun dude!
Do you have a video for testing hard drives? Also, would you happen to know how to test EMC, NetApp, and other hard drives? I have used HP and Dell servers to test the drives using their own software, but I could not figure out EMC and other drives, especially those fiber channel drives.
Yes I do and yes I have... The easiest way to do this is with a USB stick with Linux on it and use my videos to reformat and testing your disks as needed. Also don't over look your SAS controller diag tools that can diag disks on the fly plus reformat and fast int. your drives. Your RAID and HBA controllers run diags on your disks at level one and level two and will fail a drive if they fail those post BIOS diag checks. The only thing that the controllers will not do is the level 3 HDD RAM cache test. That is the cache on the drive and that will require a bandwidth testing tool from Crystal reports HDD tool set but that is all you need. Check out my video on reformating a disk using Linux from sector 520 (EMC and NetApp formats) over to PC satandard for PC and x86 servers as 512 sector formatting. Also check out my video on how to do this type of reformating on scale with many disks at once to speed things up. Hope this helps! And enjoy your fun!
What software are you using to thoroughly test and reinitialize the disk surfaces of spinning disks in an array?
Well first your HBA or RAID Controller run a series of DIAG tests for pass of fail and that is key and fastest to do. next you can use the basic DMS (Disk Management System) tools of windows or Linux to do the basic testing and formatting. Lastly you can use the Crystal disk reporting tools to some level. You see these disks (Most of them) are enterprise level disks and run in a different format. It is not like a PC hard drive. Those PC based disk tools will work but you have to place all of your enterprise drives as RAID 0 or stand alone mode and that is a lot of work. SO recommend the tools that I use. Start with your HBA or RAID controller first to see if you have an issue or not and these controllers are really good at doing this testing. Then work from there. Hope this helps some... Keep enjoying this stuff as you learn more
On PC I've used hdsentinal's surface restore function (write and check AA, 55, AA, 00 to every byte on the disk). Each sector is measured for response time after resurfacing so identify any weak sectors after surface restore. Then it reports the disks overall health. So, not relying solely on S.M.A.R.T. quick or long scan to tell me the overall health of the drive. The process can take a looooong time, like days.
I was hoping you had discovered an automated way do something similar with a disk array setup and hot swap bays. Like plug in one or more drives. System automatically queues it up to be scanned/restored. As each surface restore is completed it prints out a health report label on an attached label printer. Then each morning I look at the labels printed out for the drives that have completed, remove them, apply the label, put in another drive, etc. No need to type in any commands during any of the drive scanning.
Whats your experience with CX2 ssd or the adata su650? The prices of these ssd have dropped to a level that i can consider them to replace my 900gb sas.
I have used them both. I use the CX2 1TB and 2 TB with my disk arrays and the su650 as well but mostly for system boot drives for you can remove them from the 2.5 casing and the SATA SSD board is really small. Fun to work with them that way but my main say is the CX2 when I buy in bulk SSD drives. Hope that helps some.
Are you saying I shouldn't be storing my ram stick in a dollar store plastic tub??🙃
LOLOLOL WWwwwwEEEeeeeLLLLLLLLL....
hey I use a food grade tub with silica gel packets for my excess drives so moisture doesn't kill them and if a HDD in my NAS dies I just go in and look for a stack with the same storage size if none then the array may gain more space.
16:00 Why isn't that just standard for SSDs? the plastic does nothing but make it look nice and I am all about performance over looks.
Yea... Well...
I thought only Brendan Gregg was allowed to yell at drives
Very true!