November 4, 20169 yr 12gb in both my servers, do the first is getting upgraded to 32 because i use it as server and gamer machine for my Girl friend
November 5, 20169 yr 20G. I started with 4G. Since I started experimenting with VM and Dockers, I thought it more sense to put in some more memory.
November 23, 20169 yr Using 8GB since 7 years - useless. The sys takes only 1.06GB so i will reduce to 4GB.
November 24, 20169 yr Using 8GB since 7 years - useless. The sys takes only 1.06GB so i will reduce to 4GB. And how do you plan to reduce? Remove one of the DIMMs, crippling it to single channel? Or actually swap both out for 2x2GB? You are aware that it will be using all of the memory available to it as cache, right?
November 24, 20169 yr Using 8GB since 7 years - useless. The sys takes only 1.06GB so i will reduce to 4GB. And how do you plan to reduce? Remove one of the DIMMs, crippling it to single channel? Or actually swap both out for 2x2GB? You are aware that it will be using all of the memory available to it as cache, right? Swaped both out and used 2x2GB. This isnt a problem in my case because unRAID is only working as Plex-Server and therefore, 2x2GB are ok. EDIT: Ok, 4GB is crap - all things are very slow - back to 12GB
December 8, 20169 yr 96GB ECC. Why? Because I am planning to use it to test auto provisioning using cloud platform.
December 8, 20169 yr I can't believe people still reply to this thread. It was a poll that closed on May 21, 2014. (Two and a half years ago) Everyone just replying on the thread saying how much they use, isn't really giving interesting /useful data as its not being displayed anywhere expect just reading through alot of threads. So @limetech why not open a new poll here , as it seems still quite popular. Then we can see the info as a bar chart and get an idea of the current use in peoples servers?
December 9, 20169 yr Just my opinion, but I'm not sure a new poll would be much better. If you look back through them, it's really hard not to think they're largely ego driven, clearly not representative. If you've splurged on RAM, or otherwise proud of your RAM purchase, you're more likely to want to mention it. A new poll with the right verbiage could be better, but how useful would it be? A good general recommendation based on extent of VM usage would be more useful, I think. What would be nice, would fill a big hole, is a hardware registry of some sort, collected with user effort, and without personally identifying info. It's proven too hard to maintain a Hardware Recommendations list. It would probably require a special piece of software to both collect the info, and provide access, plus a common persistent storage location. Just ideas - * each major hardware component, with a user grade, plus a clue about how heavy the system usage is: - motherboard, CPU(s), RAM, hard drives, solid state drives, graphics cards, drive controllers, power supply; plus the VM count and Docker container count - user grade of component: --- A : completely satisfactory, even for future expansion if user wanted to --- B : completely satisfactory --- C : it's OK, adequate --- D : works, but not recommended --- F : trash, run away * current system status: - Active - Failed - Replaced - Not using * Optional user ID This would allow others to explore what motherboards are currently being used, and what components work with them. But this will never work without widespread buy in. It would have to be easy and almost automatic. Perhaps if the collection was automatic and tied into something we almost all used, like unRAID itself (means LimeTech would have to develop it) or something Squid has done like Community Apps. Most of the info can easily be collected from the syslog and lspci etc (not grades unfortunately). But there would be enough to be very helpful.
December 9, 20169 yr Perhaps if the collection was automatic and tied into something we almost all used, like unRAID itself (means LimeTech would have to develop it) or something Squid has done like Community Apps. Most of the info can easily be collected from the syslog and lspci etc (not grades unfortunately). But there would be enough to be very helpful. lol And get back into the whole "calling home" debate. But regardless, as a purely opt-in feature, it would be extremely helpful in determining what works and what doesn't.
December 9, 20169 yr lol And get back into the whole "calling home" debate. But regardless, as a purely opt-in feature, it would be extremely helpful in determining what works and what doesn't. How about a button to push next to the diagnostics button? Something along the lines of "Preview upload system profile to internet", which generates the detected output to a page, along with optional ratings and short comment fields, and has a button at the bottom, "Upload Now". Maybe generate a unique hash of the USB GUID in the submission so the results can be linked to your online profile if you wish, or kept anonymous if you don't tell anyone your hash. Don't allow editing of the hardware profile or hash, just display and submit as is. If you don't want the info sent, don't upload it. Limetech would have a record of which public IP uploaded the profile, but if it's totally voluntary and disclosed, I don't see how anyone would have a serious issue with it.
December 9, 20169 yr There is already such a feature under Tools -> Hardware Profile ... So there is... But, it actually doesn't go far enough with what I believe is the intent of RobJ's posting. You'd really need to add correlation to passed through devices, and basic specs of VMs. (Under the assumption that it actually works properly) Then LT has to actually do something with the raw data and keep a web page continually updated with the basic stats gathered from all users to assist other users make purchasing decisions, etc.
December 9, 20169 yr There is already such a feature under Tools -> Hardware Profile ... So there is... But, it actually doesn't go far enough with what I believe is the intent of RobJ's posting. You'd really need to add correlation to passed through devices, and basic specs of VMs. (Under the assumption that it actually works properly) Then LT has to actually do something with the raw data and keep a web page continually updated with the basic stats gathered from all users to assist other users make purchasing decisions, etc. I was referring to the opt-in feature to share system information. You are right that additional processing is required to translate it into 'usable' information, but that is up to LT
March 9, 20179 yr 72GB in the main server with a few vm's and plex/crashplan. Main VM used for video editing. 40GB in the backup server, which also has a few vm's/plex/crashplan. 12GB in the other 2 servers which only host a single vm.
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.