Both jobs are interesting. Both jobs at at 10km from home (cycle distance). Pay is the same (Philips grade VG45) Both are too short (one is a 5 month project and one is a 6 month project)
I think the HightTechCampus/UltraSound one is going to win. It has a prelimenary start datye of 7th of January and lasts 6 months. The other starts in Feb (3 weeks later) and only lasts 5 months.
Both are for Philips Healthtech
And both offers came out of the blue.
I've been searching for 5 months. Found nothing. Then within a week I get 2 offers
Isn't that why you make sure that a phone has a replaceable battery before you buy? (Now just crashes, also to replace the battery after about 4 years)
@Hennes You realize that most phones are not designed to last more than three years?
I tend to give mobile devices a three-year service life, after which replacement is seriously considered. Most people replace their phones after two years these days.
One thing to note about music on YouTube: When people say "listen in HD", it may not make sense as HD applies to the video stream, not the audio. However, YouTube does adjust audio quality based on the video quality. h3xed.com/web-and-internet/…
The best audio quality is obtained at 720p and higher, with 192 Kbps AAC. Dropping to 480p or 360p reduces the bitrate to 128 Kbps, resulting in a slight but noticeable reduction in sound quality for many people. At 240p, the audio bitrate further drops to 64 Kbps and is encoded in MP3; this results in significant and clearly audible loss of fidelity.
I evaluated the cost of running a server at home for large amounts of data, with a build plan similar to what I discussed many months ago. At $1700 for the hard drives alone (8x4 TB WD Red Pro in RAID 6, 24 TB usable), plus $700-$1200 for the motherboard, processor (Xeon E3 v5), and other parts, plus cost of electricity (about $13-15 a month at 100 W) means it is just not worth it compared to cloud storage.
$1200 for all the other parts, not to mention the cost of replacing failed drives (I expect one or two drives to fail over three years' time).
CapEx of about $3000, plus OpEx of about $700-$1000 over three years.
Storing 20 TB of data in Backblaze B2 costs $100 a month, or $3600 over three years—and it takes much less work to maintain. For less data, the break-even time is even longer.
It's only a good idea if I need access to data quickly or constantly stream data from the server. Backblaze B2 charges $0.05 per GB of data downloaded.
For archival storage, a home NAS is not worth the time and expense.
$3700 TCO over three years (plus the need for manual maintenance) makes no sense for archival or backup storage.
Hence, absent media streaming or other frequent data access, a home NAS does not make sense from a TCO standpoint.
Besides, for a good portion of the time, it's likely there will be much less than 20 TB of data stored, which means that the cloud option will cost significantly less in practice than the $3600 over three years.
I've decided that a NAS makes no sense for backup/archival use only. Cloud storage is more resilient and costs less.
The intended use case is backup archiving of critical data like photos.
Media streaming has not been a significant use case for me and is not likely to arise in the future. Even if it does become necessary, I might as well just plug in an external hard drive into my router and it'll appear as a network share.
Extrapolating from these computations, I'd have to store at least 16-24 TB of data in order to have a shot at breaking even within three years. Cloud is the way to go.
Even at the lowest cost (5x Seagate Archive 8 TB drives in RAID 6, 24 TB usable), the drives will cost about $1100.
At least one more drive for replacement in the event of failure will need to be kept on hand.
Downgrade processor and RAM: about $150 saved.
Power cost will be a bit lower, too (probably -$200 over three years).
That's... still about $2500-$3000.
Again, with cloud storage, you only pay for what you use, so I suspect that real cost will be about one third to one half of the $3600 over three years, which assumes 20 TB of data from the start. In practice, usage will be a lot lower than this in the beginning and grows over time.
Hence, absent the need for quick access to data or very large amounts of data stored, cloud storage makes more sense.
There are other ways for me to quickly stream or otherwise consume the data if necessary that are a lot cheaper.
Even if I'm being conservative at $2000 for 20 TB of storage over three years, cloud still makes more sense. Again, you only pay for what you use with cloud services, so unless you have 20 TB of data from the start (which is definitely not going to be the case), a storage server will wind up costing more.
It's only a good idea if I need access to data quickly or constantly stream data from the server. Backblaze B2 charges $0.05 per GB of data downloaded.
Ah well. I guess TCO is not vastly better in the cloud, but it does invalidate my NAS build plans, having found that it's not cost-effective for archival use except with extremely large amount of data.
@bwDraco There's always tradeoffs, and in this case it's probably mostly the network bandwidth and time required. I'm not saying you definitely shouldn't use cloud storage. I'm just objecting to your local storage cost analysis.
Mostly cause I think it's way overkill.
$600 for 3x 8 TB HDDs. $500-$550 for a (used) LTO-5 drive + SAS controller. $300 for ~24 GB worth of LTO-5 tapes.
You don't even need a separate computer/CPU/etc., assuming you have a decently modern desktop.
"cloud" primarily gets you some more redundancy, at the cost of depending on the company not collapsing and also the previously mentioned network transfer issues.
It's not cheaper than the (minimal) home setup, though. Not by a long shot.
In fact, if you're only doing bulk backups/archives that you won't access often, you could ditch the HDDs entirely and just double up on the tape storage.
(Of course, tapes themselves are kinda finicky to use...)
