[tor-relays] Questions about Tor consensus weight & swag

Hello,

so my relay regulary gets overloaded, for what I can only assume is an
hardware issue, (which I can't upgrade rn) since I already applied the
tor anti ddos scripts. It seems to be able to recover, however with a
lower Consensus Weight. I also (stupidly) tried to have a cron restarting
my tor daemon daily which also resulted in the latter. So I wonder, if
there is any way to have a relay run more stable and I suppose with a
somewhat higher consensus weight (I can only asssume making some further
changes in advertising bandwidth etc.

My second question(s) is/are concerning the tor swag (I hope that's
allowed to ask here). Firstly, what is the actual Speed requirement? In
the tor ecosystem, every unit is MBs, and only on the swag site its KBs.
But if my calculations are correct, 500KBs = 500000Bs = 0.5MBs which
doesn't really make sence, imo(but I probably misscalculated somewhere).
Secondly, does running mean it's uptime (aka Last Restarted) or the Time
it was first seen?

Sorry for the wall of text for not much information, and thanks a lot in
advance for your replies!

Best,
shruub

···

_______________________________________________
tor-relays mailing list
tor-relays@lists.torproject.org
https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-relays

Hi,

You shouldn't worry too much about your relay sometime appearing as
overloaded. Sadly due to the ongoing network ddos, many relays show up
that way. You shouldn't restart your relay daily, that would make everybody
think it's not very stable, so you won't be able to get flags such as Stable,
Guard or HSDir. In general to get a higher consensus weight, you need to
have higher throughput, if your relay sometimes shows as overloaded, there
is probably limited headroom for improvement.

I think the speed requirement was decided a very long time ago and wasn't
ever updated. It's 0.5MByte/s (=4Mbps) for non exit relays, and half as much
for exit ones. I guess not too many people are asking for tshirts, so having it
that low isn't an issue for the people handling that. Running means first time,
little over a dozen relays have not been restarted in the last 6 months. In
general you should try to keep your relay be up to date, which necessarily
imply a restart once in a while.

Best,
trinity-1686a

···

On Sat, 18 Feb 2023 at 11:43, shruub via tor-relays <tor-relays@lists.torproject.org> wrote:

Hello,

so my relay regulary gets overloaded, for what I can only assume is an
hardware issue, (which I can't upgrade rn) since I already applied the
tor anti ddos scripts. It seems to be able to recover, however with a
lower Consensus Weight. I also (stupidly) tried to have a cron restarting
my tor daemon daily which also resulted in the latter. So I wonder, if
there is any way to have a relay run more stable and I suppose with a
somewhat higher consensus weight (I can only asssume making some further
changes in advertising bandwidth etc.

My second question(s) is/are concerning the tor swag (I hope that's
allowed to ask here). Firstly, what is the actual Speed requirement? In
the tor ecosystem, every unit is MBs, and only on the swag site its KBs.
But if my calculations are correct, 500KBs = 500000Bs = 0.5MBs which
doesn't really make sence, imo(but I probably misscalculated somewhere).
Secondly, does running mean it's uptime (aka Last Restarted) or the Time
it was first seen?

Sorry for the wall of text for not much information, and thanks a lot in
advance for your replies!

Best,
shruub

_______________________________________________
tor-relays mailing list
tor-relays@lists.torproject.org
tor-relays Info Page

_______________________________________________
tor-relays mailing list
tor-relays@lists.torproject.org
https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-relays

I also (stupidly) tried to have a cron restarting

my tor daemon daily which also resulted in the latter. So I wonder, if
there is any way to have a relay run more stable and I suppose with a
somewhat higher consensus weight (I can only asssume making some further
changes in advertising bandwidth etc.

Restarting it on demand is the worst thing you can do. Even if it's overloaded, it's much better to leave it running as overloaded and have stable uptime history than to restart it. Restarting it makes you lose flags, and it also might be the reason your consensus weight was lowered.

If you know the limits of your resources you are better of by using:

RelayBandwidthRate X MBytes
RelayBandwidthBurst Y MBytes
MaxAdvertisedBandwidth Z MBytes

(I prefer just using the first two without MaxAdvertisedBandwidth).

It's usually a good idea to have RelayBandwidthBurst much bigger than RelayBandwidthRate. Example X = 8 Y = 20, you will see constantly 8 MB/s via your machine. If that is OK for your CPU, RAM and bandwidth, otherwise use other values.

If you don't have enough bandwidth to overload your CPU/RAM you don't need to set this as Bandwidth authorities will assign your relay such a low weight that it won't reach the bottleneck.

My second question(s) is/are concerning the tor swag (I hope that's
allowed to ask here). Firstly, what is the actual Speed requirement? In
the tor ecosystem, every unit is MBs, and only on the swag site its KBs.
But if my calculations are correct, 500KBs = 500000Bs = 0.5MBs which
doesn't really make sence, imo(but I probably misscalculated somewhere).
Secondly, does running mean it's uptime (aka Last Restarted) or the Time
it was first seen?

You can set MBytes, GBytes, KBytes - don't need to compute yourself the Bs in order to reach a desired value of MBs.

As for what is the minimum required speed, there are mixt opinions here. There is _certainly_ a threshold where a relay becomes a problem rather than useful, when the resources spend to measure it, include it in the consensus and deliver its descriptors to clients overweight the speed it provides. There's not a value established yet for this as far as I know, but it certainly should be one. 500 KBs is too little for our days, IMO, at least 1 MB/s is even too little for mobile devices on 3G (we have 5G ready...).

You mean the `Running` flag? That means the relay is `running` $now where $now is the last time the majority of directory authorities voted that they can reach your relay (usually, last authority vote).

you also have `last restarted` and `first seen` separately for the other values you mentioned. `first seen` is only used for metrics, historic purposes while `last restarted` has some effect over flags (`HSDir`, `Guard`, `Stable`, etc.).

Thanks for running a relay!

···

shruub via tor-relays wrote:

Hi,

You shouldn't worry too much about your relay sometime appearing as
overloaded. Sadly due to the ongoing network ddos, many relays show up
that way. You shouldn't restart your relay daily, that would make
everybody
think it's not very stable, so you won't be able to get flags such as
Stable,
Guard or HSDir. In general to get a higher consensus weight, you need
to
have higher throughput, if your relay sometimes shows as overloaded,
there
is probably limited headroom for improvement.

I think the speed requirement was decided a very long time ago and
wasn't
ever updated. It's 0.5MByte/s (=4Mbps) for non exit relays, and half as
much
for exit ones. I guess not too many people are asking for tshirts, so
having it
that low isn't an issue for the people handling that. Running means
first time,
little over a dozen relays have not been restarted in the last 6
months. In
general you should try to keep your relay be up to date, which
necessarily
imply a restart once in a while.

Best,
trinity-1686a

Hello,

so my relay regulary gets overloaded, for what I can only assume is an
hardware issue, (which I can't upgrade rn) since I already applied the
tor anti ddos scripts. It seems to be able to recover, however with a
lower Consensus Weight. I also (stupidly) tried to have a cron
restarting
my tor daemon daily which also resulted in the latter. So I wonder, if
there is any way to have a relay run more stable and I suppose with a
somewhat higher consensus weight (I can only asssume making some
further
changes in advertising bandwidth etc.

My second question(s) is/are concerning the tor swag (I hope that's
allowed to ask here). Firstly, what is the actual Speed requirement?
In
the tor ecosystem, every unit is MBs, and only on the swag site its
KBs.
But if my calculations are correct, 500KBs = 500000Bs = 0.5MBs which
doesn't really make sence, imo(but I probably misscalculated
somewhere).
Secondly, does running mean it's uptime (aka Last Restarted) or the
Time
it was first seen?

Sorry for the wall of text for not much information, and thanks a lot
in
advance for your replies!

Best,
shruub

_______________________________________________
tor-relays mailing list
tor-relays@lists.torproject.org
tor-relays Info Page

_______________________________________________
tor-relays mailing list
tor-relays@lists.torproject.org
tor-relays Info Page
You shouldn't restart your relay daily, that would make everybody

think it's not very stable, so you won't be able to get flags such as
Stable,
Guard or HSDir.
Ironically, I've got all of these flags, might have gottem them
afterwards I suppose.

It's 0.5MByte/s (=4Mbps)

Alright. It's interesting that this hasn't been abused (ig the tor
community are just nice people :D)

Anyways, thanks for your reply!

···

Feb 18, 2023 12:31:31 trinity pointard 'trinity.pointard at gmail.com' <obsvj67j+tor-relays=lists.torproject.org@4wrd.cc>:

On Sat, 18 Feb 2023 at 11:43, shruub via tor-relays > <tor-relays@lists.torproject.org> wrote:

_______________________________________________
tor-relays mailing list
tor-relays@lists.torproject.org
https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-relays

makes you lose

flags, and it also might be the reason your consensus weight was lowered.

I'm not sure if I actually lost flags, but yes, that is probably the main
reason.

RelayBandwidthBurst

One question, what actually is the burst? Haven't found anything online
nor in man.

There's not a value established yet for this as far as I know,

but it certainly should be one. 500 KBs is too little for our days, IMO,
at least 1 MB/s is even too little for mobile devices on 3G (we have 5G
ready...).

Indeed, yes. I'm not sure why, but every request on mobile feels a lot
slower than on a pc

You mean the `Running` flag? That means the relay is `running` $now

where $now is the last time the majority of directory authorities voted
that they can reach your relay (usually, last authority vote).
Sorta yes, basically wanted to know if it has first seen (in metrics) or
just how long it's been running.

Thanks for running a relay!

:slight_smile:

···

_______________________________________________
tor-relays mailing list
tor-relays@lists.torproject.org
https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-relays

man torrc:
RelayBandwidthBurst N bytes|KBytes|MBytes|GBytes|TBytes|KBits|MBits|GBits|
TBits
If not 0, limit the maximum token bucket size (also known as the burst) for
_relayed traffic_ to the given number of bytes in each direction.
They do not include directory fetches by the relay (from authority or other
relays), because that is considered "client" activity. (Default: 0)

https://onbasca.readthedocs.io/en/latest/bandwidth_tor.html

···

On Samstag, 18. Februar 2023 18:26:55 CET shruub via tor-relays wrote:

> RelayBandwidthBurst

One question, what actually is the burst? Haven't found anything online
nor in man.

--
╰_╯ Ciao Marco!

Debian GNU/Linux

It's free software and it gives you freedom!