new to eventstore, wondering if these log entries are normal?

EventStoreDB
1

Below are some log entries that are from one of our es clusters, are these normal?

October 9th 2019, 20:17:46.415
[00001,17,00:17:46.414] ELECTIONS: (V=566192) TIMED OUT! (S=ElectingLeader, M=NULL).

October 9th 2019, 20:17:46.415
[00001,17,00:17:46.414] ELECTIONS: (V=566193) VIEWCHANGE FROM [172.16.12.114:2112, {49f78233-c139-4a2b-b0ee-c68aeb0b1d01}].

October 9th 2019, 20:17:46.415
[00001,17,00:17:46.414] ELECTIONS: (V=566193) SHIFT TO LEADER ELECTION.

October 9th 2019, 20:17:45.412
[00001,17,00:17:45.411] ELECTIONS: (V=566192) VIEWCHANGE FROM [172.16.12.114:2112, {49f78233-c139-4a2b-b0ee-c68aeb0b1d01}].

October 9th 2019, 20:17:45.412
[00001,17,00:17:45.411] ELECTIONS: (V=566192) SHIFT TO LEADER ELECTION.

October 9th 2019, 20:17:45.411
[00001,17,00:17:45.410] ELECTIONS: (V=566191) TIMED OUT! (S=ElectingLeader, M=NULL).

October 9th 2019, 20:17:44.407
[00001,17,00:17:44.406] ELECTIONS: (V=566191) VIEWCHANGE FROM [172.16.12.114:2112, {49f78233-c139-4a2b-b0ee-c68aeb0b1d01}].

October 9th 2019, 20:17:44.406
[00001,17,00:17:44.406] ELECTIONS: (V=566191) SHIFT TO LEADER ELECTION.

October 9th 2019, 20:17:44.406
[00001,17,00:17:44.406] ELECTIONS: (V=566190) TIMED OUT! (S=ElectingLeader, M=NULL).

October 9th 2019, 20:17:43.406
[00001,17,00:17:43.402] ELECTIONS: (V=566190) VIEWCHANGE FROM [172.16.12.114:2112, {49f78233-c139-4a2b-b0ee-c68aeb0b1d01}].

October 9th 2019, 20:17:43.406
[00001,17,00:17:43.402] ELECTIONS: (V=566189) TIMED OUT! (S=ElectingLeader, M=NULL).

October 9th 2019, 20:17:43.406
[00001,17,00:17:43.402] ELECTIONS: (V=566190) SHIFT TO LEADER ELECTION.

2

Some more that are slightly different:

[00001,296,23:59:04.201] Dropping HTTP send message due to TTL being over. "ViewChangeProof" To : 172.16.13.125:2112


Dropping HTTP send message due to TTL being over. "SendGossip" To : 172.16.8.74:2112


[00001,23,23:58:46.740] Looks like node [172.16.8.74:2112] is DEAD (Gossip send failed).
3

This looks like the nodes can’t talk to each other (timeouts). This is trying to elect a master. Given v > 500,000 I am guessing there have been a lot of these…

4

Yeah, this is running in kubernetes (GKE), which from what I’ve been reading isn’t all that stable or recommended.

5

This looks more like they aren’t able to talk to each other. This isn’t “oh an election occurred and it shifted, then another 12 minutes later”. This is there have been 500,000 attempts.


Virus-free. www.avast.com

6

Cluster status shows that they are indeed able to communicate, but there have definitely been some issues this the crazy amount of attempts to elect.