Howdy ya'll!
I'm an engineer from Texas. Communications, Internet technologies, and *nix engineering are my specialties. I'm a complete noob, with regard to FreeNAS. However I started with FreeBSD, as a replacement for the commercial BSDI I used extensively. I'm not sure the year, '95 or '96 perhaps. Back then we were building terminal servers for our dialup modem banks.
My first NAS box was based on a FreeBSD beta that had experimental ZFS support... I built a large array of cheap drives for a backup system, when tapes just couldn't keep up. I've also got a Nexenta NAS serving Iscsi to a couple of VMware servers... That has been rock solid, but it is on enterprise hardware, spinning some rather expensive drives. I've been eyeing FreeNAS ever since I finished it.
So this isn't my first Rodeo.
In my day job I work with Internet technologies for a company that makes commercial products for the large service providers. So I work with a lot of the same technologies in a commercial environment, that I am here to experience with FreeNAS. In my free time, I do things like automate the bbq pit with a beaglebone and put it on the Internet, so that I can watch the temperature of my brisket while I run to the store for beer.
My FreeNAS build will be based on re-using a $1500 computer case I have owned since 2000. It is a large server case with drawers for 2 maximum-size motherboards, so it can hold nearly all form factors. It boasts support for up to 8 standard power supplies, 36 5.25" hard drives, and was incredibly heavy before I stripped it down to gut it of old parts, to make room for the new. So it has lots of nice room for things like hot-swap drive backplanes...
I'm thinking that one drawer will hold a FreeNAS system, and the other will be a Mythbuntu system. Normally I would just build custom systems, based on FreeBSD and Ubuntu, however I like the idea of tracking these open source projects, as they will do most of the hard work for me. After all, I'm not getting paid to build this, this is for relaxation... Nothing more relaxing than letting someone else do your engineering heavy-lifting. I get to just enjoy. The only downside to a project like this is that the plugins might be a couple revisions older than what I could have in a manually installed FreeBSD system, which is not a problem unless there is a new feature you want. But most package management systems for linux distributions have the same problem, so I'm cutting slack for that.
For my first taste of FreeNAS, I simply stacked some drives on top of one of my linux systems, move drive cables around temporarily, and booted off a USB flash stick. There I got an idea of how things were going to work, and fit together. No streaming video tests yet, this was just a chance to play around with the UI, and decide if FreeNAS was up to par. Seems to be!
So for the real build, I'm going with a Asrock C2750D41. I know there are mixed opinions about this motherboard. My take on this is after several years experience with custom processors for video processing. I have seen some really cheap slow machines process video better than xeon based systems. The chips I work with at work have very little horsepower, use almost no power, yet handle full resolution video at high bit rates with ease. That is what I want, low power consumption, flawless performance. Yeah, this build won't be feeding iSCSI to a VM server, all it has to do is one stream of transcoding, and perhaps a couple video streams, and some utility storage fuctions such as being the archive point for my various system backups.
Hard drives will be the most power hungry part of the system, and I *hope* I'm able to spin those down and keep the whole system in a low-power mode, when it is idle, which is a good 70% of the day. At the end of the day, the NAS is for me, and the streaming media library is for my wife. That they live in the same box is just a bonus. Mythbunto, on the other hand, is going to be my wife's machine, and my pain in the ass... I expect much more work there than with FreeNAS.
Of course, now that I'm telling people about the plan, things will change and it will all come together in a different way... So don't hold me to any of this.
Once I have a solid build running, I'll be playing with various STB's, ranging from a Roku, to raspberry pi. I expect an android based STB will be the likely favorite, once prototyping is done, and I start spending money on new STB's. I have a couple manufacturers in China constantly trying to send me some sample google STB's that look really promising, and are about 1 generation newer than the boxes for sale now on Amazon (so they will probably be buggy too...).
So try not to byte my head off in the forums for not using StuperMicro... I have my reasons. I have my FILL of stupermicro at work, and I'm not real happy about them. I mean, they are on-par with Dell, and have some really nice rack servers on paper... But out of the box they are nowhere near as solid as my real enterprise class machines. I'm sure Stupermicro is better than Asrock for many reasons... But I see some reasons Asrock will be better for ME.
I'm an engineer from Texas. Communications, Internet technologies, and *nix engineering are my specialties. I'm a complete noob, with regard to FreeNAS. However I started with FreeBSD, as a replacement for the commercial BSDI I used extensively. I'm not sure the year, '95 or '96 perhaps. Back then we were building terminal servers for our dialup modem banks.
My first NAS box was based on a FreeBSD beta that had experimental ZFS support... I built a large array of cheap drives for a backup system, when tapes just couldn't keep up. I've also got a Nexenta NAS serving Iscsi to a couple of VMware servers... That has been rock solid, but it is on enterprise hardware, spinning some rather expensive drives. I've been eyeing FreeNAS ever since I finished it.
So this isn't my first Rodeo.
In my day job I work with Internet technologies for a company that makes commercial products for the large service providers. So I work with a lot of the same technologies in a commercial environment, that I am here to experience with FreeNAS. In my free time, I do things like automate the bbq pit with a beaglebone and put it on the Internet, so that I can watch the temperature of my brisket while I run to the store for beer.
My FreeNAS build will be based on re-using a $1500 computer case I have owned since 2000. It is a large server case with drawers for 2 maximum-size motherboards, so it can hold nearly all form factors. It boasts support for up to 8 standard power supplies, 36 5.25" hard drives, and was incredibly heavy before I stripped it down to gut it of old parts, to make room for the new. So it has lots of nice room for things like hot-swap drive backplanes...
I'm thinking that one drawer will hold a FreeNAS system, and the other will be a Mythbuntu system. Normally I would just build custom systems, based on FreeBSD and Ubuntu, however I like the idea of tracking these open source projects, as they will do most of the hard work for me. After all, I'm not getting paid to build this, this is for relaxation... Nothing more relaxing than letting someone else do your engineering heavy-lifting. I get to just enjoy. The only downside to a project like this is that the plugins might be a couple revisions older than what I could have in a manually installed FreeBSD system, which is not a problem unless there is a new feature you want. But most package management systems for linux distributions have the same problem, so I'm cutting slack for that.
For my first taste of FreeNAS, I simply stacked some drives on top of one of my linux systems, move drive cables around temporarily, and booted off a USB flash stick. There I got an idea of how things were going to work, and fit together. No streaming video tests yet, this was just a chance to play around with the UI, and decide if FreeNAS was up to par. Seems to be!
So for the real build, I'm going with a Asrock C2750D41. I know there are mixed opinions about this motherboard. My take on this is after several years experience with custom processors for video processing. I have seen some really cheap slow machines process video better than xeon based systems. The chips I work with at work have very little horsepower, use almost no power, yet handle full resolution video at high bit rates with ease. That is what I want, low power consumption, flawless performance. Yeah, this build won't be feeding iSCSI to a VM server, all it has to do is one stream of transcoding, and perhaps a couple video streams, and some utility storage fuctions such as being the archive point for my various system backups.
Hard drives will be the most power hungry part of the system, and I *hope* I'm able to spin those down and keep the whole system in a low-power mode, when it is idle, which is a good 70% of the day. At the end of the day, the NAS is for me, and the streaming media library is for my wife. That they live in the same box is just a bonus. Mythbunto, on the other hand, is going to be my wife's machine, and my pain in the ass... I expect much more work there than with FreeNAS.
Of course, now that I'm telling people about the plan, things will change and it will all come together in a different way... So don't hold me to any of this.
Once I have a solid build running, I'll be playing with various STB's, ranging from a Roku, to raspberry pi. I expect an android based STB will be the likely favorite, once prototyping is done, and I start spending money on new STB's. I have a couple manufacturers in China constantly trying to send me some sample google STB's that look really promising, and are about 1 generation newer than the boxes for sale now on Amazon (so they will probably be buggy too...).
So try not to byte my head off in the forums for not using StuperMicro... I have my reasons. I have my FILL of stupermicro at work, and I'm not real happy about them. I mean, they are on-par with Dell, and have some really nice rack servers on paper... But out of the box they are nowhere near as solid as my real enterprise class machines. I'm sure Stupermicro is better than Asrock for many reasons... But I see some reasons Asrock will be better for ME.