Search results

I can not imagine huge VM overwriting completely the whole disc once per day, every day during 5 years...

Fortunately, memory-management in linux works in much more sophisticated way...

It is actually very different. If you boot live-disc, you can mount some static libs/bins from it, but otherwise you have to keep the whole filesystem in RAM. And in case of PVE it would not be so easy, as it is type-2 hypervisor (on top of full OS), so you could count some GB for it if you want...

I do not understand how exactly you'd like to modify it. What should kernel do when VM reached its limit of ram+swap usage, and yet some process wants to allocate more memory?

There are a few high-endurance USB-3.0 sticks too, i.e. Mach Extreme MX-ES or Winkom Pendrive. Given that these use SLC-NANDs, they can hold at least as long as any consumer-SSD. With price ~1.5€/GB they are not that expensive...

Slightly OT, but for those looking for possibility to have PVE+Docker on single host, there was interesting article on ServeTheHome recently. From what I understand, it does not run Docker-engine on LXC/VM but on host, in addition to KVM/LXC. The only drawback is having separate GUIs for...

I have read that wiki-page a few times and yet I'm not sure I understand it correctly. It seems to me to be a little bit simplified, because in case of using ZFS for VM-data on PVE-host, there are at least 3 different disk-related software-caches: system-cache on host (free -m), ARC on host, and...

Then why using swap at all? The last time I have been using swap on my servers was some 5 or 6 years ago. ddr price is so low it is not problem to buy extra ram if needed...

Check "free" output. BTW, when your node is using swap, it needs more ram, not more swap...

smartmontools do not go well with lsi (or any) raid controllers, even with jbod. Somewhere on avago/lsi web I have seen recommendation not to use it. Even on smartmontools-web you can find a notice it works with *some* raid controllers, and does not work with others... The other point is, there...

Have you followed load on PVE-server? Might be it ran out of memory too...

Seems like serious problem to me. Before doing anything else, I would try to save all user-data I can (from within VM, when as you say VM does not report any problem)...

Server will no go over free RAM, because disk-cache is "low-priority" usage of RAM. If any process needs more RAM then there is free at that moment, cache size will be reduced. That's the way every modern OS works (yes, even windows)... I trully do not understand this obsession with free RAM...

It is a pitty open-zfs does not include native encryption, with all zfs-like goodies (per dataset/volume/file, etc), while zfs for solaris does. AFAIK, this feature has been requested on ZoL mailing-list many times...

Apparently, different applications understand free memory differently. But all apps should clearly distinguish between "free" and "available" memory. Problem is some older versions of "free" command do not show available memory (they show +/-buffers/cache instead). That's confusing...

If there is once some contest for "the most frequently asked question", this one might win it! Honestly, did you ever use "search" function, here or google? OK, forget it. Just point your browser here, read it and check your last picture again!

Then give at least ~30GB to zfs. When you have that much RAM, use it. And any using of RAM is better than not using it at all. You could also give more to VMs, i.e. for /tmp in tmpfs (ram-disk)...

What I did is I gave ZFS as much RAM as I could while my PVE is guaranteed NOT to use swap at all (even if all VMs are running and using all vRAM they have). In my case it is 10GB (out of 64GB), while my storage-size is ~1TB (SSD-only)...

Intel 3600/3610/3700/3710 is marketed as "high endurance technology" MLC. Similar in performance is Samsung SM863-series, with "3d-MLC". Both of them have power-protection, 5 or 10 years warranty, and TBW at least a few PB. If you need higher capacity, go SAS-way, that is Micron S610/630/650DC...

Using the same logic we could say MLC is "just" variant of SLC. Then all SSDs are SLC, right? I do not think so. There is big difference between SLC/MLC/TLC NANDs, and I would never recommend using SSD with TLC-NANDs for PVE. Just search this forum for posts concerning problems with "health"...