[rs] Remote Scheduling v1

af123 said:
There's no reason that a web server needs to respond to ping requests. I've asked the ISP if they can enable it though since they allow it for IPv4.
Click to expand...

Thanks. As I said in my previous reply, disabling pings on IPv6 really is not a good idea. I know we've got used to disabling them on IPv4 for network load and denial of service attack reasons. But IPv6 really doesn't work well without pings, they've been much extended in scope in IPv6.
 
Thanks for the info. I have yet to get my head around IPv6 (or more particularly what practical difference it will make to me).
 
Owen Smith said:
Thanks. As I said in my previous reply, disabling pings on IPv6 really is not a good idea. I know we've got used to disabling them on IPv4 for network load and denial of service attack reasons. But IPv6 really doesn't work well without pings, they've been much extended in scope in IPv6.
Click to expand...
Disabling ICMP isn't a good idea under IPv6, but then it wasn't under IPv4 either although you could get away with it without breaking most things apart from PMTU discovery. Disabling ICMP code 8 (ping), as the ISP has done, should not make a difference to your connectivity though. They have said they will enable ping but in the meantime you can ping ping.citrus-it.net if you want to test RTT. Haven't tried it with IPv6 myself as I'm out and about but I'll be on a native IPv6 connection on a customer site later today. They have a couple of Gb/s to the Internet so I'll see what the performance of RS is like too.
 
Black Hole said:
Thanks for the info. I have yet to get my head around IPv6 (or more particularly what practical difference it will make to me).
Click to expand...

There are certain sites which can only be reached over IPv6. Regions that were late to the Internet party have nothing like enough IPv4 addresses to give out, Asia being particularly short. In China home users get an IPv6 only connection. At some stage (and I don't know when) IPv4 will only reach a subset of the web sites you want to visit.

The other thing with IPv6 is there is no NAT or port mappings. Every device has a real external address on the worldwide internet, so the only barrier to connection for gaming or torrents etc. is the firewall. At the moment NAT and port mappings and the application sharing features in home routers are all one big bodge which aren't properly designed or standardised across manufacturers, and as a result some things don't work. Network gaming is probably the biggest sufferer here, with torrent servers or other servers at home the next most affected.

If you just read email and browse the web then NAT works fine. Anything more complicated and NAT causes problems.
 
Owen Smith said:
At some stage (and I don't know when) IPv4 will only reach a subset of the web sites you want to visit.
Click to expand...
It's going to be a long time. IPv4 has been "running out" for years but they keep inventing new ways to conserve it (I'm actively involved in rolling out carrier grade NAT at ISPs). I also build IPv6 networks but take-up is still very slow.
 
af123 said:
It's going to be a long time. IPv4 has been "running out" for years but they keep inventing new ways to conserve it (I'm actively involved in rolling out carrier grade NAT at ISPs). I also build IPv6 networks but take-up is still very slow.
Click to expand...

I regard Carrier Grade NAT as a great evil, everyone on it ends up either double or triple NAT'd (depending on implementation). One of my friends is working on the networking code for Elite Dangerous, and he can get it to work for all case (local and remote multiple players) with NAT except in the case of CG NAT which is impossible to get to work under all circumstances (local, first level NAT and second level NAT addresses, some of which the software can't discover). It's other remote players within the CG NAT that you can't reach, because that address in the middle cannot be found by either the game software or the remote server. It's a nightmare.

I'm not a computer gamer. But I went to Andrews & Arnold deliberately because all customers get a static IPv4 (and no CG NAT) and a /48 block of IPv6 addresses. My torrent server has a true IPv6 address plus the NAT IPv4 static IP. I also have a domain and DNS entries for these.

I don't begrudge you installing CG NAT, we all have to work to get paid. I don't agree with everything my workplace does. I do think CG NAT is shame, we should just bite the bullet and go IPv6. It's not hard, and we've had ample time to get on with it.
 
Also you can CG NAT as much as you like, but if the remote site is IPv6 only then you can only reach it through one of the several types of gateway technology, which adds another level of complication and things to go wrong.

Several of the major US ISPs have just turned IPv6 on for home customers. Soon two thirds of home users in the US will have native IPv6. That's a large enough number of people that it becomes plausible for some services to say to hell with the IPv4 CG NAT luddites, we can't be bothered to make our service work over it.

The UK is lagging dangerously behind in IPv6, we're barely a blip in the worlwide stats. I regard this as a risk to the future UK economy.
 
I've done some testing of RS on a native IPv6 connection and it seems fine; faster than on IPv4 if anything. Here's a bit of the network traffic trace from Firebug running in Firefox. I cleared my cache before this test. It could be something related to your connection - try http://rs.4.hpkg.tv to force IPv4.

rs6.png
 
af123 said:
I've done some testing of RS on a native IPv6 connection and it seems fine; faster than on IPv4 if anything. Here's a bit of the network traffic trace from Firebug running in Firefox. I cleared my cache before this test. It could be something related to your connection - try http://rs.4.hpkg.tv to force IPv4.
Click to expand...

Thanks for the forced IPv4 address, that goes through without a delay.

I can ping google with IPv6 so I don't think the delay is at my end, but at least I now have an option without having to completely disable IPv6 so thanks for that.

I'll keep checking to see when ping starts working over the IPv6 address.
 
Hmm.. I can ping it now (2a00:5600:1600:0:c0:ffee:dc2:50). The ping you showed in an earlier post was a different IPv6 address.

Edit: ah, you were pinging hummypkg.org.uk and not rs.hummypkg.org.uk
The AAAA record for the former was wrong but shouldn't matter for RS
 
Ten minutes apart:

C:\Users\Owen Smith>ping rs.hummypkg.org.uk
Pinging hummypkg.org.uk [2a00:5600:1600::1:1] from 2001:8b0:2ff:cd73:3522:4f3c:4
0d1:e350 with 32 bytes of data:
Request timed out.
Request timed out.
Request timed out.
Request timed out.

and:

C:\Users\Owen Smith>ping rs.hummypkg.org.uk
Pinging hummypkg.org.uk [2a00:5600:1600:0:c0:ffee:dc2:50] from 2001:8b0:2ff:cd73
:3522:4f3c:40d1:e350 with 32 bytes of data:
Reply from 2a00:5600:1600:0:c0:ffee:dc2:50: time=41ms
Reply from 2a00:5600:1600:0:c0:ffee:dc2:50: time=39ms
Request timed out.
Request timed out.
Ping statistics for 2a00:5600:1600:0:c0:ffee:dc2:50:
Packets: Sent = 4, Received = 2, Lost = 2 (50% loss),
Approximate round trip times in milli-seconds:
Minimum = 39ms, Maximum = 41ms, Average = 40ms

Looks like it was giving me the wrong address for rs.hummypkg.org.uk. Note that it always says I'm pinging hummypkg.org.uk even though I asked for rs.hummypkg.org.uk. Are the rerverse lookup entries correct?
 
Yes - just checked, rs.hpkg.tv is a CNAME for hpkg.tv so you were being caught by that wrong record. Should be fine now (give or take any propagation delays)
 
mihaid said:
After clicking bbc1hd option in first picture you get the second picture but the record buttons have now disappeared which makes it useless. disregard signature, I'm uptodate.
Click to expand...
Fixed, thanks for the report.
 
Can I ask that the "Add New Entry" button is placed at the top and the bottom of the page as scrolling back down to add a new entry can be a real pain.
 
I was planning to add wake/sleep timers to have my HDR Fox T2 wake up at xx:45 every hour and sleep again at xx:48 every hour. Being 180 seconds long that should be enough for the 60 to 90 seconds after boot to kick in for the remote scheduling software. The timeframe was chosen to avoid waking up for padding before a recording (at about xx:53) and avoid my 04:20 to 0:4:40 reminder which prevents OTA.

But I can'd do this or anything close to it because the RS portal will only allow wake up and sleep timers to be set in 15 minute intervals from a drop down menu. Could manual editing of the times be added please so that I can set any wake up and sleep time?
 
Most wear and tear on any mechanical device occurs when power cycled.

I would certainly not want any of my HDD power cycling every hour.

The HDD is better left spinning than starting and stopping.

That is just my opinion anyway.
 
Wallace said:
Most wear and tear on any mechanical device occurs when power cycled.

I would certainly not want any of my HDD power cycling every hour.

The HDD is better left spinning than starting and stopping.

That is just my opinion anyway.
Click to expand...

Unfortunately with RS we have no way of leaving the hard disc spinning.
 
You must log in or register to reply here.
Back
Top