Live data from Hacker News

LackRack: IKEA's cheapest table is perfectly sized to rackmount computers (2020)

wiki.eth0.nl

91–100 of 250 posts

Re: LackRack: IKEA's cheapest table is perfectly sized to rackmount computers (2020)

#91

Given how many Lack tables I've seen fall apart without any sort of load, I would be very reluctant to try this, no matter how many people report it works for them. It's IKEA's cheapest table for a reason.

In my experience, even cheap IKEA stuff is pretty solid and well worth the cost as long as you treat it with care. Some stuff is garbage even if you're careful but I've never personally seen something from IKEA that isn't great if handled gently.

The IVAR system is, in my book, absolute complete junk. I couldn't get it to even attempt to work correctly, it wobbled when it wasn't collapsing, and to make insult even worse, it was more expensive than actual industrial wire shelving.

Re: LackRack: IKEA's cheapest table is perfectly sized to rackmount computers (2020)

#92

Earlier quoted context omitted.

I think this is aimed at cheapo home-labbers like myself. I doubt enterprises would consider putting a £20,000 rendering machine inside one of these.

Another option might be stereo rack mount equipment. Probably cheaper than enterprise, and usually designed to be attractive.

Stereo (like telecom) racks may not have sufficient airflow, especially when "closed" - have to check.

The best racks are the surplus ones left laying around; if you know someone who deals with buildings after companies have moved out you'll have an endless supply of racks.

Re: LackRack: IKEA's cheapest table is perfectly sized to rackmount computers (2020)

#93

Earlier quoted context omitted.

S3... Harder to use... Gotta pay egress.... It seems worse all round.

On top of that, Imgur automatically deletes EXIF data, which would've prevented GGP from leaking their home address.

I am not a machine.

Re: LackRack: IKEA's cheapest table is perfectly sized to rackmount computers (2020)

#94
post #91

Earlier quoted context omitted.

In my experience, even cheap IKEA stuff is pretty solid and well worth the cost as long as you treat it with care. Some stuff is garbage even if you're careful but I've never personally seen something from IKEA that isn't great if handled gently.

The IVAR system is, in my book, absolute complete junk. I couldn't get it to even attempt to work correctly, it wobbled when it wasn't collapsing, and to make insult even worse, it was more expensive than actual industrial wire shelving.

Did you forget to add cross-braces [0]? Ivar has always been rock-solid for me.

[0] https://www.ikea.com/us/en/p/observatoer-cross-brace-galvani...

Re: LackRack: IKEA's cheapest table is perfectly sized to rackmount computers (2020)

#95

You know you are in HN when people use S3 to upload photos, and not imgur.

If you’ve already got a bucket ready to roll for this sort of thing, S3 is cheap, straightforward, with a stable interface and API, ownership over your assets, and a much lower chance your content will end up placed on-page and strongly associated with stuff that isn’t yours.

Re: LackRack: IKEA's cheapest table is perfectly sized to rackmount computers (2020)

#96

Earlier quoted context omitted.

S3... Harder to use... Gotta pay egress.... It seems worse all round.

On top of that, Imgur automatically deletes EXIF data, which would've prevented GGP from leaking their home address.

oh ya rookie move - stripped that and wiped my CF cache

Re: LackRack: IKEA's cheapest table is perfectly sized to rackmount computers (2020)

#97
post #89
post #66

Earlier quoted context omitted.

This reminds of the many recipes that convert from Fahrenheit to Celsius (and vice versa for that matter) and don't round off the numbers properly. I'm not setting my oven to 176 degrees Celsius even if I could.

This non-rounded numbers make people think that there's something terribly important at being X degrees and not X+1, not realizing that the internal temps vary widely even when set specific. I feel recipes have caused more people to avoid cooking than anything invented in the last 1000 years, heh.

With regard to recipes, I sometimes wonder if "round numbers" are causing us to miss out on more optimal recipes. Maybe that dish really would turn out better cooked at 176F, or with 1 and 7/16 cups of flour instead of 1.5.

Re: LackRack: IKEA's cheapest table is perfectly sized to rackmount computers (2020)

#98
post #87

I naively put an old LaCie 1U server under a Lack table 13 years ago, thinking it would be a great media server in my living room. I had no idea how loud rackmount fans were. It was obviously unusable for watching any kind of media.

Modern rack servers at least spin down the fans when the system isn't heavily loaded, but it's still way too loud for a media room. For those playing at home, buying a 2U or even 4U rack will on average be quieter (still sounds like a jet engine just after reboot until the SMC comes online).

> even 4U rack

Supermicro builds some workstations[0] which were quite quiet back in Xeon 5xxx days.

[0] https://www.supermicro.com/en/products/x11/systems/workstati... be sure to check for the 'quiet fans', not the 'heavy duty' ones!

Re: LackRack: IKEA's cheapest table is perfectly sized to rackmount computers (2020)

#99

can somebody with more server/home server experience explain to me the benefit of going with a rackmount setup vs let's say, just building a second PC to use as a home server? I recently bought a house and am excited to get into home networking/home server work, but don't know anything about rack mount. Is rackmount equipment cheaper than regular consumer PC hardware?

The only real reason to rack mount at home is organization: If you have a firewall, switch, router, a server, a NAS, and a backup battery, it's really nice and clean to enclose it all in a rack. However, rack equipment will always cost more than non-rack equipment because rack equipment isn't targeted for consumers. An APC UPS that fits in a rack is $500, when the equivalent standalone one is $150. Noise is another d…

> when the equivalent standalone one is $150.

There often aren't actual equivalents available in consumer, so it's hard to get an apples-to-apples comparison. Which isn't to say a consumer grade piece of kit won't be fine for your use, just that you aren't paying $500 vs $150 for "the same thing", rather the $500 one has a bunch of feature and/or component quality stuff you may not care about (and some extra margin, but nothing like 200% you suggest)

Post reply on HN