Files are on a server (Hyper-V in this case). Quite literally on each desktop you navigate to the UNC path and do a right click on the main exe and choose send to desktop.
It seems fast when they are closed, but slow when they are open (and everyone is using it). They only have 9 PC's however. The server is way above min and even recommended spec's for the software.
Their network switch is a Netgear GS748Tv5 and I upgraded the FW to the latest version.
@black3dynamite I should have by now, but no. Let me give that a shot.
Got it! I have tried both. No noticeable change.
The way the app is installed is that essentially all the server is doing is hosting a file share. You place a shortcut to the .exe file on the desktop.
Even if it doesn't matter. I have run the powershell command and it's still running slow. Any other ideas?
I have unteamed, and disabled VMQ on the Hyper-V settings. I haven't disabled it on the NIC yet. No change so far.
I assume the powershell command will take them offline for a moment... Do I need to have the VM's shut down first before running it?
Yes, LAN drivers are up to date. I'll try the VMQ and also unteaming.
Nothing looks glaringly wrong?
This is Server 2016 and Intel... I hadn't heard of anything but broadcom with the vmq issue.
Going to try that this weekend, but wanted to have a game plan 1st.
I have 2 Intel GB NIC's Teamed together for 2 GB. I have a Hyper-V network adapter that is used for 2 VM's. VM1 - DC/DNS/DHCP, VM2-Application server
I have some thin clients and they RDS into the apps server as well as some traditional desktops that just run the application (a veterinary software that is a flat file database) over the network.
I have set this up 30-40 times before but only 1 other time using teaming.
The application is fast on the RDS session but very slow over the network. The previous server (a 10 year old server that was entry level) was running faster than the new server that has 64 GB Ram, 10K RPM Drives in RAID 10 when running over the network.
Am I not setting this up correctly?
Rather than place, how about format?
Does anyone have a nice template for keeping track of IP's, Passwords, Network info, etc.
I am using Google Docs right now (thinking about going WIKI) but format is the greatest issue I am having to deal with.
@scottalanmiller said in RAID on SSD's:
@ccwtech said in RAID on SSD's:
@scottalanmiller said in RAID on SSD's:
That's not a way to measure RAID reliability. I actually a video about why we never mention "number of lost drives" in the SAMIT queue.
Which one? You have several RAID videos.
It's in the queue, not visible yet. It's specifically about why we don't look at "how many drives can be lost"
Ahhh, ok.
@scottalanmiller said in RAID on SSD's:
That's not a way to measure RAID reliability. I actually a video about why we never mention "number of lost drives" in the SAMIT queue.
Which one? You have several RAID videos.
@scottalanmiller said in RAID on SSD's:
@ccwtech said in RAID on SSD's:
Would a RAID 6 be better for the extra redundancy?
RAID isn't that simple. Does the cost for extra parity make sense?
Was looking at the fact that you can lose 2 drives (not just one) as an extra safeguard. But in a 4 drive SSD RAID 10 makes more sense to me than 6.
So for greater redundancy RAID 10?
Is the reason for 5 vs. 10 just the cost of the extra drive?
@travisdh1 said in RAID on SSD's:
@ccwtech said in RAID on SSD's:
For some reason I recall that booting to a RAID 5 was unreliable (shouldn't do), but it's been so long since I have worked with RAID 5 I'm not sure where I heard that.
RAID 5 on HDD is unreliable today, and shouldn't ever be used. SSD drives don't have the same issues. That might be what you're remembering.
I believe so as well.
For some reason I recall that booting to a RAID 5 was unreliable (shouldn't do), but it's been so long since I have worked with RAID 5 I'm not sure where I heard that.
