Missed Block Leader 0

Staking & Delegation Operate a Stake Pool
1

According to CNCLI I had a block assigned this epoch 266 epoch_slot_no: 120320.
I am processing transactions, my op.cert was just renewed several times in the past weeks, Tip Diff is in good shape and I have between 10-15 inbound peers connections.
Is not even that the block was not adopted, LiveView shows the following:

│ BLOCKS                            Leader | Adopted | Invalid │
│Since node start (EKG metrics)     0      | 0       | 0       │

Usually if is a problem on the pool side it should show Leader: 1 and 0 adopted.
My slot appears to be empty on db-sync.

I am not sure how to read this, what is most likely gone wrong in this case?

thanks

2

Slot is missing from Cardanoscan

3

yes, what does that mean? something is wrong on my side?

4

Do you have any missing slots with cardano_node_metrics_slotsMissedNum_int ?

Have you installed chrony?

5

chrony is installed and running as daemon… I had a clock issue then?

$ sudo chronyd -q
2021-05-17T10:44:32Z System clock wrong by -0.000226 seconds (step)

cardano_node_metrics_slotsMissedNum_int 307

6

I don’t think it is chrony, but you need to monitor the missed slots.

307 in how many days?
Can you check to see when those missed slots happened?

Is it a local Producer bare metal or virtual? How much ram does it have?

For reference:
2021-05-17T11:11:39Z System clock wrong by -0.000040 seconds (step)

I have 0 missed slots since the epoch start and I had 0 missed before the epoch switch an uptime in total of a week.

7

right, I see, I guess I don’t have enough knowledge to investigate this.
I am on Bare Metal ram=32Gb
As I don’t have any delegators was waiting to bring things on virtual so I currently run 1 Relay and BP-Node on the same machine

8

You have enough ram.
Maybe the problem is running both the relay and BP on the same machine.

If you can transfer your relay to cloud or another machine might solve your missing slots issue.

9

I had similar issue two month ago… if you dont have metrics about Leader event, it means the node was not scheduled. hope only cncli calculated wrongly the schedule…

10

is this common?

11

I may try that, is there anything I can do to investigate the cause of those 307 slot missing?

12

mempool trace enabled in your cfg?

13

Use prometheus and grafana!

Then you can track all the metrics and check what else is happening at the time of a missing slot ie CPU spike, Memory dumped, etc

1 Like