Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
 
  • Post
  • Reply
Munkeymon
Aug 14, 2003

Motherfucker's got an
armor-piercing crowbar! Rigoddamndicu𝜆ous.



GnarlyCharlie4u posted:

Plex is great, although I've never exposed mine to the internet. I'm aware of split tunneling though I've never implemented it.
It would work if I were at like Starbucks or something, but the whole VPN from inside my office network, and installing the client software, is apparently a no-no. So the whole VPN route might just be dead in the water anyway.

I bought the Plex Pass and it Just Works with no VPN fuckery. Plex is gr8

Adbot
ADBOT LOVES YOU

BurgerQuest
Mar 17, 2009

by Jeffrey of YOSPOS

Munkeymon posted:

I bought the Plex Pass and it Just Works with no VPN fuckery. Plex is gr8

^ this

abelwingnut
Dec 23, 2002


looking for some advice on my current nas setup.

currently i have a 4-bay synology configured for shr-2. i have a 3tb, 2tb, and 2 1tb drives in there, giving me 2tb of available space, 2tb of protected space, and 3tb of unused space.

i would like to change this, i think. i'm starting to get to a point where this isn't enough. furthermore, i feel shr-2 may have been overkill for a home system. am i generally correct in that?

so i'd like to improve on both of those things--i want more space, and i want to change the configuration.

i know i can't convert an shr-2 to shr-1 directly. i think what i would have to do is buy higher-capacity drives to the point of getting enough unused space so that i could create a different shr-1 storage pool, move the contents of the original shr-2 pool to the new storage pool, delete the old storage pool, and then consolidate the new unused space into the new pool. is that the best way to attack this?

theoretically, that makes sense. but i'm trying to do something similar now with the 3tb of unused space i have now and the option is greyed out. i'm guessing because it isn't possible with the drives i have.

i don't know, it's all getting confusing to me at this point. i'm looking to optimize what i have for as little as possible. i wouldn't mind just replacing all the drives with 4 4tb drives or 6tb drives if the price makes sense. 4 4tb drives would give me 8tb in shr-2 and 12tb in shr-1. that latter option is very attractive, but i just don't know if i can switch to that given the current setup.

thoughts? am i missing something obvious?

also, what drives are you all buying? i have wd reds. someone in another thread said those are a bit off these days and i want to avoid low part numbers. but i'd be buying on amazon or newegg or something like that and i don't know how i could possibly check on that.

don't know, just generally throwing my situation out there in hopes of bettering it. thanks for listening.

KOTEX GOD OF BLOOD
Jul 7, 2012

Smashing Link posted:

Well it turns out Synology is continuing to act like Good Guys and replacing my 1515+ that was a few months out of warranty. It is the 2nd one they have replaced for me.
What's been going wrong with them? I have three Synology units and they all have been working great with zero problems for years.

DrDork
Dec 29, 2003
commanding officer of the Army of Dorkness

KOTEX GOD OF BLOOD posted:

What's been going wrong with them? I have three Synology units and they all have been working great with zero problems for years.

There's a hardware bug in the Intel Atom C2xxx chips that will eventually render it unbootable. The Synology xx15 line used a lot of those chips, so eventually they will fail to come back up from a boot.

Rooted Vegetable
Jun 1, 2002

Buff Hardback posted:

The plugin Unbalance in Community Apps lets you do this sort of thing easily, by letting it scatter a drive to other drives.

Haven't had a chance to post in a few days but I wanted to concur. Unbalance is a furiously usefull app when the need comes up.

One of my drive is throwing errors so I've just excluded it from being used from all important/irreplacable shares and copied the data off the erroring drive. TBH for my easily replaced stuff I'll just run it into the ground.

EDIT: If all else fails, I've got Crashplan properly up to date.

Actuarial Fables
Jul 29, 2014

Taco Defender

abelwingnut posted:

currently i have a 4-bay synology configured for shr-2. i have a 3tb, 2tb, and 2 1tb drives in there, giving me 2tb of available space, 2tb of protected space, and 3tb of unused space.
...
4 4tb drives would give me 8tb in shr-2 and 12tb in shr-1. that latter option is very attractive, but i just don't know if i can switch to that given the current setup.

If you want to convert to shr-1, you'll have to copy your data off the NAS, delete the shr-2 array, create an shr-1 array, then move your data back onto it. Since you're looking to pick up higher-capacity disks that are individually bigger than your current data partition this should be feasible - copy your data to a new disk, create a shr-1 array with the three other disks, copy the data over to the new array, add the 4th/data disk to the unit and expand the array.

The WD Red 2tb, 3tb, 4tb, and 6tb drive models that end with "EFAX" are not good and should be avoided. They are even listed as incompatible on Synology's compatibility list. Other WD Red drives and the new WD Red Pro drives are still good to use, and other NAS drives from other manufacturers are fine as well, like Seagate's Ironwolf line.

A popular trend if you're looking for even higher capacity drives (8tb+) is to buy external drives and shuck them. The WD EasyStores use NAS/Enterprise drives and they're often a lot cheaper than buying the drive itself.

KOTEX GOD OF BLOOD
Jul 7, 2012

DrDork posted:

There's a hardware bug in the Intel Atom C2xxx chips that will eventually render it unbootable. The Synology xx15 line used a lot of those chips, so eventually they will fail to come back up from a boot.
Holy poo poo, uh, thanks for letting me know about this. Is it fixable in software? I'll be checking when I get home if I have one of those procs.

DrDork
Dec 29, 2003
commanding officer of the Army of Dorkness

KOTEX GOD OF BLOOD posted:

Holy poo poo, uh, thanks for letting me know about this. Is it fixable in software? I'll be checking when I get home if I have one of those procs.

Nope. Your options are to (1) never reboot/shut down an impacted device forever, (2) RMA it, (3) google for how to solder a 100ohm resister onto the motherboard that will effectively fix the issue.

Henrik Zetterberg
Dec 7, 2007

Heners_UK posted:

Haven't had a chance to post in a few days but I wanted to concur. Unbalance is a furiously usefull app when the need comes up.

One of my drive is throwing errors so I've just excluded it from being used from all important/irreplacable shares and copied the data off the erroring drive. TBH for my easily replaced stuff I'll just run it into the ground.

EDIT: If all else fails, I've got Crashplan properly up to date.

Yeah this is super cool. Thanks to the OP for mentioning it.

abelwingnut
Dec 23, 2002


Actuarial Fables posted:

If you want to convert to shr-1, you'll have to copy your data off the NAS, delete the shr-2 array, create an shr-1 array, then move your data back onto it. Since you're looking to pick up higher-capacity disks that are individually bigger than your current data partition this should be feasible - copy your data to a new disk, create a shr-1 array with the three other disks, copy the data over to the new array, add the 4th/data disk to the unit and expand the array.

The WD Red 2tb, 3tb, 4tb, and 6tb drive models that end with "EFAX" are not good and should be avoided. They are even listed as incompatible on Synology's compatibility list. Other WD Red drives and the new WD Red Pro drives are still good to use, and other NAS drives from other manufacturers are fine as well, like Seagate's Ironwolf line.

A popular trend if you're looking for even higher capacity drives (8tb+) is to buy external drives and shuck them. The WD EasyStores use NAS/Enterprise drives and they're often a lot cheaper than buying the drive itself.

thank you, thank you, thank you. will proceed per your suggestion.

e: one follow up, does rpm really matter with a nas?

abelwingnut fucked around with this message at 23:10 on Jul 17, 2020

Actuarial Fables
Jul 29, 2014

Taco Defender

abelwingnut posted:

thank you, thank you, thank you. will proceed per your suggestion.

Thinking about it now, once you have your bigger drives ready, you could remove one of your 1TB drives, place in one of your new drives, create a single-disk SHR volume, copy everything over, then replace the other disks and expand the new volume. Your current array can lose two disks and operate fine, so removing one of the disks isn't an issue, and since everything will be done on the unit itself it will probably go faster with less room for error.

As a note, I haven't done this type of data migration on a Synology before.

quote:

e: one follow up, does rpm really matter with a nas?

Faster disks have better performance, but you won't notice any benefit in a home NAS where you're limited to a gigabit connection and have only a few users accessing documents and media.

Actuarial Fables fucked around with this message at 23:27 on Jul 17, 2020

KozmoNaut
Apr 23, 2008

Happiness is a warm
Turbo Plasma Rifle




This isn't too bad, the PSU and mobo fit nicely into the Node 304 case. Let me just look at the other side...



:cthulhu::can:

I will make everything fit, with enough zip ties, everything is possible.

KozmoNaut fucked around with this message at 09:43 on Jul 18, 2020

Hadlock
Nov 9, 2004

abelwingnut posted:

looking for some advice on my current nas setup.

currently i have a 4-bay synology configured for shr-2. i have a 3tb, 2tb, and 2 1tb drives in there, giving me 2tb of available space, 2tb of protected space, and 3tb of unused space.

Buy a 6-8tb drive, and replace one of the 1tb drives. Then, next paycheck replace the other 1tb drive. Keep shr-2, gain space

I recently went from shr-1 to shr-2

Smashing Link
Jul 8, 2003

I'll keep chucking bombs at you til you fall off that ledge!
Grimey Drawer

KozmoNaut posted:



This isn't too bad, the PSU and mobo fit nicely into the Node 304 case. Let me just look at the other side...



:cthulhu::can:

I will make everything fit, with enough zip ties, everything is possible.

Nice 304. I would love to see these go on sale again.

Enos Cabell
Nov 3, 2004


Smashing Link posted:

Nice 304. I would love to see these go on sale again.

I just put together a system in a 304 this afternoon. Was fun trying to squeeze in a 1080 and full size power supply but the results are worth it.

BlankSystemDaemon
Mar 13, 2009



There's a ~$100 2.5Gbps network switch that was just released on the market.

SwissArmyDruid
Feb 14, 2014

by sebmojo
I've got a copper SFP ready to go.

edit: wait, poo poo, gently caress. I forgot, Edgerouters only have 1GB SFP unless you fork over two grand.

edit edit: ....poo poo, the EdgeMAX line is on the chopping block next, innit?

SwissArmyDruid fucked around with this message at 12:16 on Jul 19, 2020

CommieGIR
Aug 22, 2006

The blue glow is a feature, not a bug


Pillbug
Why not go straight to 10g copper? Seems like very few NICs are going to be compatible with that.

DrDork
Dec 29, 2003
commanding officer of the Army of Dorkness

CommieGIR posted:

Why not go straight to 10g copper? Seems like very few NICs are going to be compatible with that.

Probably because you can't get 10Gb copper switches for anything that would even resemble a reasonable price. Best you can do is shove copper <-> SPF adapters into SPF 10Gb switches, which works but is a very niche case because at that point why wouldn't you just use fiber in the first place instead of paying $30-$40 per transceiver?

If he's got a SFP port on whatever server he uses, he could slap a copper transceiver in there and connect it up to this 2.5Gb switch and then connect to whatever other copper devices he's got--2.5Gb is starting to appear on high-end motherboards, and it'd be a lot simpler from a home networking topology perspective than dealing with a mixed fiber/copper network. Obviously the trade-off here is the simplicity of a single physical network vs being limited to 2.5Gb instead of 10Gb.

BlankSystemDaemon
Mar 13, 2009



CommieGIR posted:

Why not go straight to 10g copper? Seems like very few NICs are going to be compatible with that.
Basically, it's all about compatibility with existing wiring. 2.5GbE can be done on 100m cat5e S/SFP, and if you've got shorter runs you can even do 5GbE - which will go 100m if you use cat6a (which is S/STP-only, ie. no UTP).

BlankSystemDaemon fucked around with this message at 16:43 on Jul 19, 2020

Saukkis
May 16, 2003

Unless I'm on the inside curve pointing straight at oncoming traffic the high beams stay on and I laugh at your puny protest flashes.
I am Most Important Man. Most Important Man in the World.
I have also gotten the feeling that 10Gbase-T is a dead end technology. The higher latency and power use will give the victory to SFP+. That has happened in our datacenters. We used to have a mix of 10G SFP+ optical and Base-T, but future installations will be SFP+. When the world was switching to 1G datacenters and homes used basically the same technology, it just needed time to become cheaper and reach homes. But if the only users for 10Gbase-T are the hardcore NAS or homelab builders, then the use case will be such a niche it will never reach economy of scale and wide acceptance. On the other hand, SFP+ ports are way too long, motherboards don't have room for it so integrated SFP+ can't become common. Not to mention the standard is too complicated compared to base-T. I'm a server admin and regularly buy them or spend a lot of time in the datacenter, and I have really started to understand the options with SFP+ in the past few years.

BlankSystemDaemon
Mar 13, 2009



Saukkis posted:

I have also gotten the feeling that 10Gbase-T is a dead end technology. The higher latency and power use will give the victory to SFP+. That has happened in our datacenters. We used to have a mix of 10G SFP+ optical and Base-T, but future installations will be SFP+. When the world was switching to 1G datacenters and homes used basically the same technology, it just needed time to become cheaper and reach homes. But if the only users for 10Gbase-T are the hardcore NAS or homelab builders, then the use case will be such a niche it will never reach economy of scale and wide acceptance. On the other hand, SFP+ ports are way too long, motherboards don't have room for it so integrated SFP+ can't become common. Not to mention the standard is too complicated compared to base-T. I'm a server admin and regularly buy them or spend a lot of time in the datacenter, and I have really started to understand the options with SFP+ in the past few years.
Yeah, 10GbE will lead to a fiber-rich future, because going from 10GbE to 25GbE and 40GbE and 100GbE is simply not possible with copper.

DrDork
Dec 29, 2003
commanding officer of the Army of Dorkness

Saukkis posted:

I'm a server admin and regularly buy them or spend a lot of time in the datacenter, and I have really started to understand the options with SFP+ in the past few years.

Yeah, the whole "lol SFPs can't auto-negotiate down to slower speeds in general" is a pain in the rear end and would probably need to get addressed before it'd be something that would be viable outside the datacenter/motivated home users.

I do wonder how much of the SFP physical format is actually required technologically vs what it's been just because.

DrDork fucked around with this message at 18:34 on Jul 19, 2020

CopperHound
Feb 14, 2012

D. Ebdrup posted:

... which will go 100m if you use cat6a (which is S/STP-only, ie. no UTP).
UTP cat 6a absolutely exists, but it is almost as big as old 25pair cables.

H110Hawk
Dec 28, 2006

DrDork posted:

Yeah, the whole "lol SFPs can't auto-negotiate down to slower speeds in general" is a pain in the rear end and would probably need to get addressed before it'd be something that would be viable outside the datacenter/motivated home users.

You're an eeprom flash away from that. I've got some qsfp's with the wrong firmware on them right now that are 10/40 instead of 25/100. Allegedly putting the right eeprom on them will fix it.

Mind you those are 4 port breakout dac's.

Ironically our servers have 2x10g copper on them we don't use. Either way dac's are the most common way to hook up in a DC and the world is done with 1/10/40/100 and on to 2.5/5/25/50/100. I'm not sorry that home use 10g never came to reality but 2.5 multiplier stuff is coming in hot and heavy.

DrDork
Dec 29, 2003
commanding officer of the Army of Dorkness
Yeah, I just mean you can't have a 10Gb SFP+ that will talk to a 1Gb SFP automatically in a lot of cases without extra steps. With copper you just jam them in and they'll talk and figure their poo poo out (with the exception of apparently 2.5/5 with some switches). No worrying about multi vs single mode, 10Gb vs 1Gb vs whatever.

H110Hawk
Dec 28, 2006

DrDork posted:

Yeah, I just mean you can't have a 10Gb SFP+ that will talk to a 1Gb SFP automatically in a lot of cases without extra steps. With copper you just jam them in and they'll talk and figure their poo poo out (with the exception of apparently 2.5/5 with some switches). No worrying about multi vs single mode, 10Gb vs 1Gb vs whatever.

Networks with lasers on their heads.

Trirate (10/100/1000) only came about because of necessity. Things moved very slowly back then and equipment was super expensive. Negotiation is for chumps these days, and once you include the assumption of light you have to make certain that both ends can see what you're shooting.

I imagine once we get to 2.5 gig on cheap Dell desktops dual rate switches will come out.

That being said apparently our 25g Intel nics fired right up once we set them to 10/40 instead of 25/100. Monday we will find out if eeprom flashing the dac's will get us to the 4x25/100 we paid for.

DrDork
Dec 29, 2003
commanding officer of the Army of Dorkness

H110Hawk posted:

I imagine once we get to 2.5 gig on cheap Dell desktops dual rate switches will come out.

True, but it seems like this may be a long time in coming: the hardware itself isn't cheap, datacenters are never going to bother with it, loving up economies of scale, 1Gb is "fast enough" for the vast, vast majority of home users right now, etc. It's existed for years but no one really seems to have much interest in it.

Honestly, we'll probably get wide-spread 10Gb for home users via wifi before we do via copper. Wifi 6 already gets real close in theory, so maybe Wifi 7 will get there in actual practice.

H110Hawk
Dec 28, 2006

DrDork posted:

True, but it seems like this may be a long time in coming: the hardware itself isn't cheap, datacenters are never going to bother with it, loving up economies of scale, 1Gb is "fast enough" for the vast, vast majority of home users right now, etc. It's existed for years but no one really seems to have much interest in it.

Honestly, we'll probably get wide-spread 10Gb for home users via wifi before we do via copper. Wifi 6 already gets real close in theory, so maybe Wifi 7 will get there in actual practice.

BlankSystemDaemon
Mar 13, 2009



I need this, SO MUCH.

DrDork
Dec 29, 2003
commanding officer of the Army of Dorkness
Where...where did you get that?

H110Hawk
Dec 28, 2006
https://www.effinbirds.com/products/gently caress-wifi-fireside-poster?_pos=1&_sid=9f85cb5ad&_ss=r

And yes, that is my zoom backdrop at work because I am a mature adult.

SwissArmyDruid
Feb 14, 2014

by sebmojo
General, broad-usage-case wifi? Yes, gently caress that poo poo.

necrobobsledder
Mar 21, 2005
Lay down your soul to the gods rock 'n roll
Nap Ghost
I'm going to wind up building conduit into the house I build in the coming years but perhaps it'll all be a nuclear wasteland in the upcoming corporate-race war anyway and I shouldn't bother.

SwissArmyDruid posted:

I've got a copper SFP ready to go.

edit: wait, poo poo, gently caress. I forgot, Edgerouters only have 1GB SFP unless you fork over two grand.

edit edit: ....poo poo, the EdgeMAX line is on the chopping block next, innit?
There's several cheaper switches with SFP+ and 10gb capabilities. The real WTF for me is the one Ubiquiti switch I saw that had SFP+ interfaces... that supported 1 gbps max.

DrDork
Dec 29, 2003
commanding officer of the Army of Dorkness

necrobobsledder posted:

There's several cheaper switches with SFP+ and 10gb capabilities. The real WTF for me is the one Ubiquiti switch I saw that had SFP+ interfaces... that supported 1 gbps max.

Cheaper than $100? Do tell! Mikrotik dropping a 4-port SFP+ 10Gb switch for $130 a bit ago was a "big deal" precisely because it was the cheapest option by far for small layout 10Gb networking. Most of the other cheapish switches are things like 8+2 where 8 ports are 1Gb and then 2 are 10Gb, which is better than nothing but still not the same.

Moey
Oct 22, 2010

I LIKE TO MOVE IT
I'd almost rather go used enterprise gear. A Juniper EX-3300 would fit my current home upgrade path.

necrobobsledder
Mar 21, 2005
Lay down your soul to the gods rock 'n roll
Nap Ghost

DrDork posted:

Cheaper than $100? Do tell! Mikrotik dropping a 4-port SFP+ 10Gb switch for $130 a bit ago was a "big deal" precisely because it was the cheapest option by far for small layout 10Gb networking. Most of the other cheapish switches are things like 8+2 where 8 ports are 1Gb and then 2 are 10Gb, which is better than nothing but still not the same.
The original question was asking around $2k which is way higher than the $300-$400-ish UniFi SFP+ devices I saw. Their UniFi switches with just SFP in them still run $150+ anyway so it makes no sense they'd drop down to ~$200 for SFP+ unless they did a huge crunch and upgraded everything to SFP+. Ubiquiti is in the process of a wi-fi generation upgrade it seems (they're starting to send out beta product releases for 802.11ax now across several SKUs). The 16 XG is pretty overkill with 12 SFP+ ports and 4 10G RJ45 ports but is probably what I'd go for with serious intent of a 10+ year prosumer home network timeline.

TenementFunster
Feb 20, 2003

The Cooler King
is there a way to stop my x570-based system from disconnecting from my DS920+ when the computer goes into sleep mode?

waking the thing up to a dozen "we lost connection to the network drive!" notifications gets old

Adbot
ADBOT LOVES YOU

BlankSystemDaemon
Mar 13, 2009



There's no way to persist TCP connections like that, as a persistent TCP connection typically requires some sort of noop to confirm that the connection still works.

Depending on the OS, you may have a way of starting things on-resume like FreeBSDs /etc/rc.resume does.

  • 1
  • 2
  • 3
  • 4
  • 5
  • Post
  • Reply