BonzaiDuck
Lifer
More than a year ago, I was sweating over "very occasional" instability on my sig-rig. When I started the trouble-shooting, I had an SSD with ISRT caching a large HDD; a second HDD was used for (mostly) DVR captures. It was "time" though -- time to buy a 500GB SSD for a boot-system disk. That system currently has the SSD system disk, a second SSD of similar size, and the HDD.
I built a second but similar system with a smaller SSD boot disk, a 60GB SSD caching a larger HDD.
My instability problems vanished a year ago with the first system, and I'm not completely sure how I fixed it.
Recently, fretting over the Magician 4.6 software and the EVO firmware upgrade, my attention was drawn to the OS-optimization screen of Magician, and thence -- to "Virtual Memory" or the page file configuration.
I may not have been thinking clearly. I reduced the page-file size for the boot-system SSDs to vary between 1K MB and 2K MB. For some reason, I didn't specify a page file for the HDDs.
This was the only significant thing I did in recent weeks. Suddenly, today -- my sig rig threw up a "freeze" -- instability for the first time in a year of more-or-less 24/7 operation (with occasional manual/deliberate reboots here and there). I noticed some Event-Viewer entries which MAY have related to a page-file-too-small. And I realized that I'd left no page-file for anything but the boot-SSD -- and a small one at that.
Researching again to refresh my brain about Windows Page-files, I saw that page-file size could affect stability.
So I set my HDDs in each system to "system managed" page-file size while leaving the boot SSDs to a maximum 2GB page-file.
The advice about page-files I churned up today suggested "don't do anything with it" unless you "have to:" Let Windows manage page-files -- make them "system managed."
Does anyone have any alternate recommendations for my current configuration(s) [boot-SSD 2GB; HDD "system managed" --- total = ~ 17GB all page-files]?
I'm also eyeballing my server: 4x 2TB HDDs (pooled) with 120GB SSD boot disk (not part of the pool). The SSD currently provides the entire cache of 8GB for the server.
I built a second but similar system with a smaller SSD boot disk, a 60GB SSD caching a larger HDD.
My instability problems vanished a year ago with the first system, and I'm not completely sure how I fixed it.
Recently, fretting over the Magician 4.6 software and the EVO firmware upgrade, my attention was drawn to the OS-optimization screen of Magician, and thence -- to "Virtual Memory" or the page file configuration.
I may not have been thinking clearly. I reduced the page-file size for the boot-system SSDs to vary between 1K MB and 2K MB. For some reason, I didn't specify a page file for the HDDs.
This was the only significant thing I did in recent weeks. Suddenly, today -- my sig rig threw up a "freeze" -- instability for the first time in a year of more-or-less 24/7 operation (with occasional manual/deliberate reboots here and there). I noticed some Event-Viewer entries which MAY have related to a page-file-too-small. And I realized that I'd left no page-file for anything but the boot-SSD -- and a small one at that.
Researching again to refresh my brain about Windows Page-files, I saw that page-file size could affect stability.
So I set my HDDs in each system to "system managed" page-file size while leaving the boot SSDs to a maximum 2GB page-file.
The advice about page-files I churned up today suggested "don't do anything with it" unless you "have to:" Let Windows manage page-files -- make them "system managed."
Does anyone have any alternate recommendations for my current configuration(s) [boot-SSD 2GB; HDD "system managed" --- total = ~ 17GB all page-files]?
I'm also eyeballing my server: 4x 2TB HDDs (pooled) with 120GB SSD boot disk (not part of the pool). The SSD currently provides the entire cache of 8GB for the server.
