|
H110Hawk posted:I would not expose a synology directly to the internet for unsolicited inbound traffic. The USA is just as nasty as the rest of the world. Ok. Just so I understand, if the VPN server is running on the Synology, that service would be exposed to the internet but then all other services can be somehow routed through that service so they are not directly exposed?
|
# ? Oct 24, 2018 17:54 |
|
|
# ? May 23, 2024 20:07 |
|
You just defined "VPN," yes.
|
# ? Oct 24, 2018 18:06 |
|
Atomizer posted:You just defined "VPN," yes. Ok thanks
|
# ? Oct 24, 2018 18:20 |
|
Smashing Link posted:Ok. Just so I understand, if the VPN server is running on the Synology, that service would be exposed to the internet but then all other services can be somehow routed through that service so they are not directly exposed? Correct. There would be (basically) 1 port open to the internet - whatever is correct to make the VPN function. Your vpn will build a tunnel to the other synology, also with only that one port open, using the public ips. These are the IPs you are using today. Then, you will only communicate over the new RFC1918-addressed network. (Think: 10.0.0.0/8, 192.168.0.0/16, 172.16.0.0/12) If you want you can just "bridge" this onto your internal network and both synology's will show up as being right there, but I would try to just let them communicate directly and you only use the one local to you.
|
# ? Oct 24, 2018 18:21 |
|
H110Hawk posted:Correct. There would be (basically) 1 port open to the internet - whatever is correct to make the VPN function. Your vpn will build a tunnel to the other synology, also with only that one port open, using the public ips. These are the IPs you are using today. Then, you will only communicate over the new RFC1918-addressed network. (Think: 10.0.0.0/8, 192.168.0.0/16, 172.16.0.0/12) If you want you can just "bridge" this onto your internal network and both synology's will show up as being right there, but I would try to just let them communicate directly and you only use the one local to you. That makes sense to me. I think I can figure it out from there. Thanks guys.
|
# ? Oct 24, 2018 18:37 |
|
Another option would be to allow access to your Synology only from the IP used by the Synology at your work. This would effectively still block it from the rest of the internet. But I suspect the Google router doesn't support this.
|
# ? Oct 25, 2018 19:41 |
|
You can potentially setup iptables based firewalls on the Synology on either and get pretty far. It mostly gets annoying when you have to deal with dynamic IPs or corporate IT blocking certain ports / protocols (and rightfully so, you could be exfiltrating company data, wtf)
|
# ? Oct 25, 2018 22:42 |
|
I can now sorta see why some people opt for the NSC-810A instead of the 810. It was quite a bear to work with getting the the motherboard situated, but uh, it's in there and it boots, so gently caress it. All that's left is taking it home, moving my drives and Unraid flash drive over and slapping my LSI card in there. Sorta wish the PSU cables were a smidge longer getting to the motherboard. I might end up getting some extension cables for piece of mind. 8-bit Miniboss fucked around with this message at 03:07 on Oct 26, 2018 |
# ? Oct 26, 2018 03:04 |
|
My Freenas server threw a smart error but when I run both a short and long test everything seems ok. The only off thing appears to be errors 1-3 of the short test but everything I've read doesn't mention any of this as indicative of anything. I'm assuming these are what got flagged? Will they get flagged every time my scheduled self tests are run?
|
# ? Oct 26, 2018 04:48 |
|
What smart error was it?
|
# ? Oct 26, 2018 05:44 |
|
Devian666 posted:What smart error was it? In freenas it was 'Self-Test Log error count increased from 0 to 1' and the day it was flagged was on my long test schedule. I ran both a long and short test on the drive it indicated but the only thing I saw that was wrong was ID 1-3 (I said errors 1-3 incorrectly in the previous post) had high counts but nothing I'm reading seems to indicate thats a problem. Neither of the tests actually indicated an error and other than IDs 1-3 it seemed to have passed.
|
# ? Oct 26, 2018 16:16 |
|
The catch this season was pretty good, boys. Testing...(definitely not going to burn the house down) Doing it right this time, with like, actual server hardware and not crap that my neighbor used to mine bitcoins with Had to split up the testing across 2 different computers, and the one with only 2 drives attached finished much quicker (the write/read speeds decrease with more drives attached, probably saturating whatever link the USB controller is connected to). The first batch was good, so let's shuck! These are HGST air-filled DC HC320 drives that have been de-tuned to 5400 RPM. While maybe not quite as nice as the helium filled Reds people were getting from other shucking drives, for $150 an 8TB datacenter drive is pretty goddamn good. Eventually going to migrate the 6 3TB Reds that I have in my current NAS over to this enclosure (so, 6x 8TB drives + 6 3TB drives, running in separate RAIDZ2 pools). They're nearing 6 years old but still don't throw any errors, so I'll keep using them for low-value storage (trashy reality shows, etc). And yeah, yeah a CLC on a server is stupid as hell but it was just laying around and I had nothing else to use it on, so whatever.
|
# ? Oct 28, 2018 00:45 |
|
Laff, the T430 I have from work has the exact same trackpad wear pattern, it’s about 5 years old. Also the last good T series thinkpad imo. What motherboard is that btw? Fakedit: ah x11ssh, nice.
|
# ? Oct 28, 2018 00:51 |
|
t430s own (and I don't hate the 440s). I resell them all day for $400 with SSDs. Business types love the hell out of them.
|
# ? Oct 28, 2018 00:59 |
|
The T440s is acceptable if you replace the no-buttons trackpad with the yes-buttons one from the T450s.
|
# ? Oct 28, 2018 01:02 |
|
Is the buttons trackpad actually better at tracking or is it just for the buttons?
|
# ? Oct 28, 2018 01:04 |
|
taqueso posted:Is the buttons trackpad actually better at tracking or is it just for the buttons? The trackpad actually sucks rear end, use the nib, it's the only way to fly
|
# ? Oct 28, 2018 01:05 |
|
forbidden dialectics posted:The trackpad actually sucks rear end, use the nib, it's the only way to fly I can't use the nib, I also can't aim on a console, might be related.
|
# ? Oct 28, 2018 01:07 |
|
taqueso posted:Is the buttons trackpad actually better at tracking or is it just for the buttons? About the same for tracking, so yeah, pretty much just for the buttons. The buttons one also has way less travel for pressing the entire trackpad to click. forbidden dialectics posted:The trackpad actually sucks rear end, use the nib, it's the only way to fly Using the nib sucks without the buttons.
|
# ? Oct 28, 2018 01:09 |
|
I could never get the hang of the nib, I am a bad thinkpad user People at my co got upgraded to hps and everyone hates them so I am hanging on to my T430 as long as possible. It is running pretty slow on Outlook etc but that’s probably just a windows install age issue, cpu should still be fine.
|
# ? Oct 28, 2018 01:38 |
|
The clit mouse is only really for touch typists right? Anyways the trackpads are just fine compared to some of the alps Dell models. Now those suuuuck.
|
# ? Oct 28, 2018 02:49 |
|
forbidden dialectics posted:Had to split up the testing across 2 different computers, and the one with only 2 drives attached finished much quicker (the write/read speeds decrease with more drives attached, probably saturating whatever link the USB controller is connected to). The first batch was good, so let's shuck! What software are you using / what's your testing process?
|
# ? Oct 28, 2018 02:58 |
|
Atomizer posted:What software are you using / what's your testing process? Since I didn't want to have to set up a dual-boot partition with Linux (and use smartctl and badblocks), I used GSmartControl to do a short SMART test (2 minutes) before starting anything, then used h2testw to write/verify each drive (~30 hours). Then I ran the "long" SMART test from GSmartControl (~15 hours). Because it's Windows, you're only running it on the filesystem, so the last couple hundred reserved MB doesn't get tested...but I'll take those odds. Without the fans, the drives were getting up to 55C, which is way too hot. Added the fans and they held at around 38C, which is perfect.
|
# ? Oct 28, 2018 03:53 |
|
D. Ebdrup posted:I saw a presentation by a university professor at a local university, in which he extrapolated that based on the current trend of growth during the highest IPv6 adoption period, IPv6 address space would experience exhaustion before all IPv4-only equipment on the internet had been replaced with dual-stack capable equipment. Every person on earth could each have roughly 40 octillion IPv6 addresses (40,000,000,000,000,000,000,000,000,000). I think that professor is dumb.
|
# ? Oct 29, 2018 15:18 |
|
Inept posted:Every person on earth could each have roughly 40 octillion IPv6 addresses (40,000,000,000,000,000,000,000,000,000). I think that professor is dumb. Sure but if at the peak adoption more ipv4 only devices were added to the point where the percent of ipv6 capable devices went down his point still stands. Graph a curve going down and let me know when it hits the limit of the y axis.
|
# ? Oct 29, 2018 15:54 |
|
I got a DS918+ to replace the linux fileserver at my small business. So far it is working fine. I setup several volumes, each with one share. Today I put in an SSD to be a read cache. It seems like I can only use the SSD to cache one volume - is that right?
|
# ? Oct 29, 2018 18:32 |
|
taqueso posted:I got a DS918+ to replace the linux fileserver at my small business. So far it is working fine. I setup several volumes, each with one share. Today I put in an SSD to be a read cache. It seems like I can only use the SSD to cache one volume - is that right? I can't seem to find anything that would allow it to cache for a second volume.
|
# ? Oct 29, 2018 20:00 |
|
So the standard way to use this is a large volume with multiple shares?
|
# ? Oct 29, 2018 20:07 |
|
taqueso posted:I got a DS918+ to replace the linux fileserver at my small business. So far it is working fine. I setup several volumes, each with one share. Today I put in an SSD to be a read cache. It seems like I can only use the SSD to cache one volume - is that right? Correct. It works by being a pass through block map in lvm so you can only use whole devices at a time. This is a Linux thing not a synology thing.
|
# ? Oct 29, 2018 22:13 |
|
taqueso posted:So the standard way to use this is a large volume with multiple shares? I always just have one large storage volume with multiple shares as you say.
|
# ? Oct 30, 2018 03:38 |
|
I have a bunch of data locked up on a windows storage pool/storage spaces, originally created with windows server 2012 r2, on drives 2,3,4,5 If I replace the boot drive (drive 1) with a new drive, and then install win 10 on there, will I be able to access the disk/storage pool/storage spaces? I am thinking the answer is yes. Trying to migrate to temporarily access the array, to transfer the data a synology NAS unit.
|
# ? Oct 30, 2018 04:32 |
|
Hadlock posted:I have a bunch of data locked up on a windows storage pool/storage spaces, originally created with windows server 2012 r2, on drives 2,3,4,5 I'm pretty sure it will work. I think Windows sees the Storage Space data on the drives themselves, rather than having it tied to a single Windows installation. I had tried it once (with 3 x 250 GB HDDs that had little other purpose) and then disassembled the array without formatting the drives, and when I replaced one of them in a system Windows started to look for the other drives.
|
# ? Oct 30, 2018 04:58 |
|
8-bit Miniboss posted:I can now sorta see why some people opt for the NSC-810A instead of the 810. It was quite a bear to work with getting the the motherboard situated, but uh, it's in there and it boots, so gently caress it. All that's left is taking it home, moving my drives and Unraid flash drive over and slapping my LSI card in there. Sorta wish the PSU cables were a smidge longer getting to the motherboard. I might end up getting some extension cables for piece of mind. The PSU cables are actually not long enough on the 810A, you need an extension for both the CPU Aux and the 24-pin. I was going to ask you whether you routed the 24-pin cable around the bottom of the motherboard or the top but I guess that model is completely different. I'll grab some snapshots when it's finished-ish. Don't feel too bad, this is a normal part of mITX builds. There is never enough space unless it's a big case with too much space.
|
# ? Oct 30, 2018 05:31 |
|
Paul MaudDib posted:The PSU cables are actually not long enough on the 810A, you need an extension for both the CPU Aux and the 24-pin. I was going to ask you whether you routed the 24-pin cable around the bottom of the motherboard or the top but I guess that model is completely different. I'll grab some snapshots when it's finished-ish. Over the top as pictured, the opening under is just a smidge too small to fit the connector under it on account of the locking latch unless I removed the cage entirely beforehand which I don't think is possible due to the front being riveted in. Think I might also replace the CPU fan. It's a bit loud for my tastes. Just a matter of finding one that will fit it and the small clearance between the motherboard and drive cage. 8-bit Miniboss fucked around with this message at 07:41 on Oct 30, 2018 |
# ? Oct 30, 2018 07:36 |
|
No, I mean our cases are completely different and the cables route in different ways, lol. email support@u-nas.com and ask him the CPU clearance height, I've bugged him on this a couple times before. I think it's a garage operation but the guy does answer emails one a week or so. I'm currently trying to figure out what this piece goes to, if you recognize this part on your chassis let me know: Personally I threw a NH-L9i on it, and the fact that cooler doesn't fit LGA2011-3 was a significant factor in deciding to go LGA1151 with an i3 for me. No idea what you're using but it looks like it fits to me, we'll see about thermals when it's all closed up, but even with the back/PSU off the chassis on the table with no cover it was running whisper quiet. It came with a pair of Gelid Silent 12's on intake and a 80mm Gelid on exhaust. I put Silverstone fan filters on the intakes and swapped the fans out with Corsair ML120 non-Pros (the filter acts like a rubber grommet to reduce noise, no need for pros). The intake fans weren't going but the outtake fan wasn't loud at all while I was browsing through the BIOS. Looks OK so far. I think the airflow model will work - in the back, out the top right. My plan for putting a SATA-to-USB adapter on the 2.5" tray drive was a little mixed... it definitely does not fit normally, the adapter is too thick to get the drive onto the tray, and the tray positioning puts the tray too close to the fan for the length of the adapter too... but I put the SSD onto the tray backwards, so the standoffs hold it farther off the tray, then bolted the whole tray onto the inside. hackerman.jpg. Not sure it'll actually fit PCIe cards like this, clearance might be a bit too tight on the inside now, but I guess we'll cross that bridge when we come to it. I think at least one of the two slots should work (the one farther away from the tray) Be aware the black front plate has a rubbery kind of cover that does scratch/mar easily, I apparently scratched it on the table a bit at some point just getting it built. Also, some of the sata cables were threaded through each other, and I did maybe fatally kink one at one point (they are the thin low-profile kind like silverstone sells). But overall it's like 8/10 or 9/10. A few things I would do differently in hindsight (no Supermicro board lol) but it's not bad at all. Biggest gripe is... you really do need 24-pin and CPU-aux extension cables even with the "approved" power supplies, and the guy pretends you don't. You absolutely do, there is no way to route the cables even close on the 810A, at best you are like 6" short and I'm not actually sure these 8" adapters will make it either. It's not like this is weird mobo layout either... standard "aux on the top, 24-pin 90 degrees clockwise, standard ATX layout" board. Generally the same layout as the board he makes, too. Paul MaudDib fucked around with this message at 08:32 on Oct 30, 2018 |
# ? Oct 30, 2018 08:08 |
|
Paul MaudDib posted:No, I mean our cases are completely different and the cables route in different ways, lol. Ah, misread. Nope, haven't seen that part before. My accessory box only came with screws, replacement screws, some zip ties and the 2.5" drive tray. 810 comes only with the 2 Gelid Silent 12's since the mobo doesn't sit at the top. Those were promptly replaced with Noctua NF-S12A's. Might do filters but I'll try unobstructed for now. I can at least control them through an IPMI plugin through Unraid rather than klutzing around with ipmitool which is sorta neat.
|
# ? Oct 30, 2018 08:38 |
|
That part looks like a mount point for an extra SSD / 2.5” drive. Compare screw holes to see if that makes sense. I’ve got a U-NAS NSC-800 that’s a mini ITX and it seems the uATX form factor version doesn’t seem much friendlier. The front of the case is a low-mid quality rubberized plastic that can handle some torsion but the coating itself is only so resilient.
|
# ? Oct 31, 2018 15:39 |
|
The U-NAS guys passed it to their engineer, who thinks it is a mount bracket that came with the PSU. Paul MaudDib fucked around with this message at 17:08 on Oct 31, 2018 |
# ? Oct 31, 2018 17:05 |
|
priznat posted:Laff, the T430 I have from work has the exact same trackpad wear pattern, it’s about 5 years old. Also the last good T series thinkpad imo. I didn't like how grody the trackpad was getting when I figured out that it's an adhesive coating that's making it look gross, rather than any type of dirt or grime. I scraped off the coating to remove the dot texture and then sanded it with 1500 grit. On top of looking clean the trackpad feels so much better now. I know you can get replacement stickers, but not spending money is better. Oh, wait, this is the NAS thread. Uhh, my FreeNAS setup in a TS140 has been chugging along just fine for years, thanks for asking!
|
# ? Oct 31, 2018 19:14 |
|
|
# ? May 23, 2024 20:07 |
|
I heard you guys like hard drives 6x8TB in one RADIZ2 pool, 6x3TB in another. About 40 TiB usable (for now). Running ProxMox with separate LXC containers for Turnkey Fileserver, Plex, SABNzdb, Sonarr, etc. ZFS pools are handled by the host and mounted as bind points to the containers.
|
# ? Oct 31, 2018 20:06 |