PoolTool Propagation Delay -4996ms (Time Sync)

Staking & Delegation Operate a Stake Pool
1

Hi all!

I was recently contacted by Serki from Ada4Science[BOSON] the other day with a heads up about some possible issues with our pool [RELAY]. Thanks Serki! <3
Just in time too! We have finally been assigned our very first slot after 4.5 months in the game! :partying_face: So we are of course adamant to make sure that block gets minted!

One of the issues Serki pointed out was the fact that we may have a time sync issue as we did not show up in the list of pools showing propagation times on PoolTool’s network health page - Cardano PoolTool - The most comprehensive staking statistics for Cardano on the web.
You can also see on our pool’s metric page on PoolTool, our propagation delay is negative 4996ms… - Cardano PoolTool - The most comprehensive staking statistics for Cardano on the web.

We have since installed chrony on the relay (almost 3 days ago) and the BP (14 hrs ago) which has indeed fixed a time propagation issue we had. According to our grafana charts, we now have a time sync of 2-8ms on both. However we still don’t show up on the network health page, and our propagation delay remains -4996ms

Is this negative value possibly due to our single bare metal relay being on the same local network as our bare metal BP at my partner’s (MVS) premises? I understand the -4996ms chart might take 1 or 2 epochs to update, but I believe we should be showing up on the network health page as this is live.

Anyone have any suggestions for us to resolve this? Or know if this is indeed an issue?!!

1 Like
2

Do u set up the sendmytip script used to send data to pooltool on Producer?

I don’t see this metric for ur pool

All Propagation Delays Producer

I see only

All Propagation Delays Receiver

Question: when did u set up the script (sendmytip)? Also, did u set the pooltool API key inside the script, right?

3

Just spoke to our operator MVS, and he confirmed he did setup the sendmytip script with our pooltool API key in it, this was done around the 18th Sept when we migrated to bare metal for both the BP and relay.

I didn’t even realize we have a metric chart missing completely!

Something to note is when chrony got installed 3 days ago on the relay, we did see our pool pop up briefly on network health, but soon disappeared and never appeared again. Also our block height is reporting correctly on pooltool

4

Try to reset the chrony service for both nodes, perhaps it will help; the script (sendmytip) is running only on Producer right?

5

Turns out we actually used cncli sendtip. Only on BP. Is it suggested to use sendmytip instead?
Restarting chrony on both nodes didn’t fix the issue unfortunately.

6

We’ve restarted all the sendtip related services and we are now showing on some block propagation times. Would having more relays in different parts of the world help with having lower propagation times across all blocks?
If I understand correctly, network health only shows pools with less than 3000ms delay for the current block. So we aren’t showing sometimes because the current block is being minted by nodes which are far away from ours geographically

7

Could be one of the reasson; how is the internet connection?

8

Hey :slight_smile:
If I remember correctly the “All Propagation Delays Producer” chart only started showing up after our first minted block. Not sure if that is coincidence though.

Fyi we’re currently using cncli sendtip only on the BP. Also we have two relays in central Europe and seem to show up pretty much always on the network health page, so I don’t think it is necessary to have a geographical distribution over the globe.

9

Gigabit! Never has issues. Hardware is also well beyond recommended specs

Hopefully we will see this very soon then :slight_smile:

Good to know, thanks!

For an update on the issue, it seems the restarting of the sendtip related services has resolved the issue, we do occasionally show up on network health now.

Now for some digging into why we don’t have a very good propagation time to nodes located far from us! My only possible thoughts are to check the internet line speeds to other countries and make sure we get appropriate latency, add a secondary relay in a country outside of the US (which we planned on doing anyway, in South Africa where I live), and possibly play with the topology file? We attempted adding some pools from Asia to the topology file that frequently had low propagation times last night, but had no luck.

10

Did you ever fix this issue, we started 2 epochs ago and still have this
speed

11

Yo! That graph does correct itself once you’ve minted your first block

1 Like