Announcement: Onboarding for (Terra Wave 3)

Node Operators
42

A few key things I appreciate you guys doing on this reassessment:

  • Providing transparency on how the entire evaluation process worked

  • Creating three separate weightings to see how it impacted the scoring to ensure this scoring methodology was meaningfully different from shortlist A

  • Removing validators who don’t run their own infra, recently faced a slashing event, or had performance below the baseline requirements

  • Putting more emphasis on commitment to the Terra ecosystem through governance participation, seniority, public services, content creation, community building, and user-facing tools

All in all, amazing work from the Lido team on a task that is difficult given the variety of factors that must be considered!

8 Likes
43

Tier 1: 18 validators appearing in 4/3 columns, issues:

  • Solva Blockchain Solutions (CryptoCrew Validators): missed one or more seed node checks in the latest Terra Delegation Program (TDP) round and didn’t qualify
  • Orbital Command (it was in the waitlist of the shortlist A, not in the shortlist A): missed one or more seed node checks in the latest TDP round and didn’t qualify
  • Allnodes: didn’t qualify for the latest round of the TDP, due to missing one or more Bombay testnet checks

Public info confirming this: TDP 2022 R1 Snapshots - Google Sheets

Tier 2: 8 validators appearing in 2 columns:

8 validators for 5 spots left, my suggestion is to either increase the number of validators selected from 23 to 26 or, if not possible, select those appearing in shortlist A + one column, versus validators appearing in 2 columns but not in shortlist A. Also, there are some large jumps in the data, for example BridgeTower Capital ranks 8th in the second column and then it doesn’t even appear on the third column. Also, some issues of validators in this Tier 2:

  • Fresh luna: not qualified for the latest round of the TDP due to missing 1 or more seed node checks
  • SynergyNodes/Larry stakehouse: as shown on the web it seems that there is only one technical person (in the case of SynergyNodes the other person seems only involved in the community) so this would be the previously mentioned “bus factor” risk?

Do Kwon mentioned “But I feel like the decision criteria here should be driven by empirical data (previous history of uptime, slashing events, and voting correctly on oracle votes)”, we have Avg Sign 99.8% and Avg Oracle 100%.
Other variables like content creation/community building/seniority are quite subjective (maybe some were not validating until recently but already involved in the Cosmos/Terra ecosystems for years, and this may affect also the governance parameter depending on whether it was calculated as a % of voting versus the total votes).

Tier 3: 6 validators in only 1 column:

5 out of 6 of these validators don’t qualify, but Mantra Dao appears also only in one column and will qualify, while some validators from shortlist A + one column don’t qualify? This doesn’t seem right/fair.

  • Mantra Dao: failed one Bombay testnet validator check in the latest round of the TDP, however this snapshot was not considered because of Christmas, otherwise it would not had qualified

And by the way there is a typo at the time of writing, the last column should be 75% and not 25% again :+1:

4 Likes
44

No list is ever perfect, but I appreciate that you are so responsive and open to community suggestions. So great work guys!

I believe a balanced approach with a 50/50 split is a good way to go forward for this application round. A balanced approach will also provide insight on how you would like to structure future rounds and then adjust the scoring settlement accordingly.

4 Likes
45

Hello,
I know it’s hard to choose validators for onboarding properly, there will be always somebody upset, so I would like to add MY VIEW on those 2 lists:

Shortlist A - AKA Network stability - List of validators that are in most cases qualified operators. They have invested heavily into having a decentralized and independent infra (no cloud), reliable team, and stable professional level operations. This is something you cannot achieve in a month or two and takes months to make it right and for most of them, it also means sacrificing community engagement.

Shortlist B - AKA Community engagement - The list of validators, that have a decent score in terms of operations, but in a lot of cases their community engagement surpasses their stability and independence of operations.

When you look at community engagement, It’s really hard to measure that right, because is not possible to include all aspects of it.

For example, in our case (RBF Staking) we are hosting an in-person COSMOS ecosystem focused conference called GATEWAY in Prague in May - https://gateway.events/
It takes tremendous time and effort to organize an event like this, but we know, that this event will bring huge value to the whole ecosystem including Terra.

This type of community engagement is not used in community engagement rules, because obviously, it’s almost impossible to correctly measure every aspect of the engagement.

You may say that a 50% alignment score seems an obvious choice as a balance between network stability and community engagement. This works only on paper, in reality, this doesn’t work. As I mentioned before, it’s impossible to measure community engagement as objectively as network stability, thus 50% community engagement spoils the results of Shortlist B.

Tomas
Head of RBF Staking

2 Likes
46

photo_2022-04-07 12.59.47
photo_2022-04-07 12.59.471280×990 305 KB

hei guys, it seems this has gone wild in all directions :slight_smile: i just want to put this up here [even thought anyone could query the data] this is just our voting, regarding to the recent easy2stake case, the systems we rely on and use are not perfect, we started our governance engagement when we joined lido, what is interesting in the graph attached is the fact that we missed 5 governance votes, and this was, like in easy2stake case, a failover of multiple systems, systems that supposed to work, we have relied on two bots for the governance proposals to alerts us, one of hours and one 3rd party, both bots/systems failed silently for that perriod and none of us from 01node had any idea about it [we are like 6 people now] and i did not manually check for that perriod, so things like this can happen, in extraordinary situations, this post is not a justification or an excuse, i just wanted to share that things like this can happen to anyone, easy2stake or us, probably others as well, we had 2 systems we took for granted, that failed at the same time.

Have a great day guys.

4 Likes
47

My point of view is that community engagement is important too and should not be disregarded.
Frankly speaking, without our communities none of us would be where are today.

Is it possible to propose to onboard more validators for this round or are we capped at 23 validators?

1 Like
48

This is a good question. Why is there a cap of 23 validators? Would not the Terra ecosystem benefit more if the anchor stake is distributed across a large number of validators?

Btw, are current validators (who are already in the Lido program) assessed against the same requirements? It could happen that not everyone stands according to the standards.

1 Like
49

Hi Lido team,

Thanks for creating these sublists.
We are aligned with you on the second option from the shortlist B subset.

It seems rationale to us to go with the proposed second option for three reasons:

I. It is true that some companies have greater focus on technical capabilities while others put greater emphasise on community engagement. In the past, both approaches have come to fruition for the companies involved, and that is because they all bring along the know-how and also human and technical capacities. Since both approaches can succeed, a 50% - 50% approach will provide the best of both worlds.

II. Significantly promoting ecosystem alignment, without introducing significant discrepancies in performance potentially constitutes a bigger value add to the Lido ecosystem.

III. The second list sets a baseline for future collaborations and potentially helps Lido to become a better decision maker in the future.

Best regards,
Andre

2 Likes
50

Answering @John_Middlewood and @OneStar questions regarding the number of validators to be onboarded.

In Lido, we’re trying to be in contact with our validators because of many questions appearing. We’re monitoring validators’ performance, help with network upgrades coordination (especially the urgent ones), and so on. So onboarding for us is not just “whitelisting the addresses in smart contract” but all the organizational work. That’s why we limit the number of validators to be onboarded on each wave trying to find the trade-off between decentralization and keeping the standarts of maintaining the best validator set.

51

Hey, Kai. Terra has a pretty robust validators community. I am sure, that there will not be needed any efforts from Lido to coordinate network updates (we’ve seen how effectively terra migrated from col-4 to col-5) or similar activities.

The expansion of validators sets will require minimum effort to handle. This will also help to avoid discussions of what approach is fair and what is not. Especially taking into account the absence of transparency from Lido side.

I am sure you have seen tweets like this: https://twitter.com/larry0x/status/1513250678348599298?s=20&t=SYHsU0zc2rklXKVEUfVXmw

Which creates some level of distrust about the selection procedure mechanics. In this regard, a larger set and transparent metrics would definitely help.

Thank you.

52

Gov. Vote Onboarding 3

Throughout March 2022, the Lido Node Operator Sub-governance Group (LNOSG) evaluated 65 onboarding applications for inclusion into the Lido on Terra Node Operator set. As a result of community feedback and the evaluations conducted on the 28th of March, the LNOSG is presenting two possible sets of validators for the DAO’s consideration, and a third choice of “none of the above” in which case the entire onboarding process would be started over.

In short, Shortlist A is the product of the original LNSOG evaluation, and Shortlist B is the result of an added emphasis to ecosystem-participation and alignment as a result of strong Terra community and DAO member feedback.

Shortlist A: 21 shortlisted applicants, original LNOSG evaluation result.

Ratings were submitted by LNOSG evaluators, averaged per applicant, and large variances or outliers were specifically discussed during the evaluation call. The criteria we examined include:

  • General validating experience both cross-chain and on the specific chain in question.
  • Overall historic performance and recent performance on the specific chain.
  • Level of detail and effort put into application so that the LNOSG may be able to assess various questions/factors around infra setup, security, reliability, overall anti-fragility of the validator, decentralization, and geographic distribution.
  • Ecosystem participation & alignment.
Shortlist A
0Base
Allnodes Inc
AUDIT.one (Persistence Staking Pte. Ltd)
Blockdaemon
BlockNgine
BridgeTower Capital
BTC.Secure
Chainlayer
Coinhall
Cosmic Validator
Delight Labs
MissionControl
moonshot
Orbital Command
PFC
RBF Staking
RockX
SkillZ
Solva Blockchain Solutions (CryptoCrew Validators)
Stakely
Terran one

NB: Two validators were removed from the shortlist following downtime that brought their performance below the baseline criteria. SkillZ will only be added to the validators registry and delegated to if they join the active set.

Shortlist B: 21 shortlisted applicants, 2 waitlisted applicants, composite score.

Shortlist B was offered as a possible alternative to Shortlist A following Terra community and DAO feedback expressing the desire for an onboarding wave of more ecosystem-aligned applicants.

For the purpose of this shortlist, a scoring methodology was adopted to “boost” the weighting of ecosystem participation criteria in the overall applicant score. This “Ecosystem Alignment Score” focused on governance participation rates, seniority, public services (open-source development, public LCD/RPC hosting…), content creation and community building.

The LNOSG Scores and the Ecosystem Alignment Scores were given equal weightings and averaged to produce a more community-representative shortlist that still takes performance and setup into consideration.

Shortlist B
0Base
Allnodes Inc
AUDIT.one (Persistence Staking Pte. Ltd)
BlockNgine
BTC.Secure
Chainlayer
Coinhall
Delight Labs
Fresh luna
Kytzu
larry stakehouse :cut_of_meat:
MissionControl
moonshot
Orbital Command
PFC
RockX
Setten
Solva Blockchain Solutions (CryptoCrew Validators)
Stakely
SynergyNodes
Terran one
Stakebin (waitlisted)
MANTRA DAO (waitlisted)

NB: shortlisting is contingent on an affirmative response from the identified applicant that they wish to continue with onboarding. If the validator doesn’t approve its participation it will be replaced with the waitlisted one.

More information about the onboarding application windows and evaluations performed can be found here: https://research.lido.fi/t/announcement-onboarding-for-terra-wave-3/1677/4.

This vote will run for 5 days and seeks LDO holder approval of the onboarding of the aforementioned node operators for the respective protocols.

Our goal with onboarding new node operators is to improve the decentralization and resilience of Lido on Terra, as well as bringing on new organizations with strong skill sets and alignment with Lido’s values to help grow liquid staking on the Terra network.

53

These two validators were Easy 2 stake and InfStones, however validators included in shortlist B MissionControl and Solva Blockchain Solutions (CryptoCrew Validators) have currently worse performance with Avg Oracle % of 94.8% and 96.4%. Their current Oracle Commit % is 0% and 73% respectively. As Do Kwon previously said with Easy 2 stake:

54

That’s right, we’ll double-check and if those validators broke the baseline, they won’t be whitelisted, the rule stays the same as in previous cases.

3 Likes
55

Hi guys, we (CryptoCrew Validators) have pointed our oracle feeder to a more stable endpoint after it’s performance degraded over the past days. Our alertsystem caught the issue, we should be missing way less votes now. We’re excited to be considered for Lido onboarding. If there are any questions, please don’t hesitate to reach out to me or our community support, @cryptocrew_validators (tg) @crypto_crew (Twitter) or here :slight_smile:

1 Like
56

It would be very helpful also for the community to have more transparency about the significance/amount of these “community concerns” that led the LNOSG for the first time to question its evaluation process and propose a different set of candidates.
I saw a couple of people here and on twitter complaining, of course with 65 applications and 21 selected many won’t be happy, but is this a reason to question the evaluation process for the first time and delay the Lido Dao voting to come up with an alternative list? Moreover, if the evaluation process was going to be changed I think this should have been done for the next rounds, not “on the go” in the current wave 3 that was already finalized.
In short, it would be really appreciated to have a compilation document of all these “community concerns” that for the first time led the LNOSG to propose a shortlist B.

57

Hi, I’m luca from the MissionControl team.
We have had some problems with our price-server that started sending wrong data to the feeder. We didn’t notice the problem immediately because our monitoring system only checked if the oracle votes were sent but not if they were correct. We have solved the issue and developed a new monitoring system that also checks the validity of the votes. If you are interested in the code it’s open-source and available on the Github of one of our developers: terra-discord-webhook/validatorOracleAlerts.js at main · alecande11/terra-discord-webhook · GitHub .

We apologize for the inconvenience, we will do everything possible to prevent this from happening again.

58

Short update regarding the onboarding process.

The voting has ended and the chosen option is Shortlist B:

image
image672×532 12.9 KB

We’ve requested the participation confirmation and some additional information from Shortlist B applicants and going to start the actual onboarding as soon as we gather it.

Unfortunately, we have to de-list MissionControl and Solva Blockchain Solutions (CryptoCrew Validators) because of the downtime incidents that have taken place recently. Both incidents broke the baselines so we’re going the same way we went with Easy 2 Stake and InfStones and can’t onboard these validators this round.

cc: @MissionControl @ccclaimens

UPD: The deeper details of downtime incidents of the validators above.

  1. For MissionControl — the baseline is broken. From April 11 to April 14 the Mission Control oracle has been sending the same price info to the chain. So the votes are completely missed for April 12 and 13 and partially for April 11 and 14. It leads to >50h of missed oracle votes streak.
  2. For CryptoCrew — the baseline isn’t really broken. The oracle was especially unstable on April 11. But in terms of being offline in a row, the longest streak is 1h 7m. So for the whole day, the amount of totally missed oracle votes is nearly half of the total votes but there was no 6-hours streak which means that the baseline isn’t broken. So for Solva Blockchain Solutions (CryptoCrew Validators), we are going to proceed with the onboarding process.

Apologize for the confusion.

2 Likes
59

1200x675
1200x6751200×675 481 KB

We are thrilled to announce that we have now received the requested information as well as interest confirmations from all selected applicants.

Validators to be onboarded
0Base.vc
Allnodes Inc
AUDIT.one (Persistence Staking Pte. Ltd)
BTC.Secure
Chainlayer
Coinhall
Delight Labs
Fresh luna
Kytzu
larry stakehouse
MANTRA DAO
moonshot
Orbital Command
PFC
RockX
Setten
Solva Blockchain Solutions (CryptoCrew Validators)
Stakebin
Stakely
SynergyNodes
Terran one

As mentioned in the previous post, we unfortunately had to delist Mission Control, as well as BlockNgine, because of downtimes that brought their performance below the baseline criteria. StakeBin and MANTRA DAO, the two waitlisted applicants, will be onboarded instead. We invite Mission Control and BlockNgine to re-apply at the next onboarding round.

In the next few days, validators will be added to the whitelist and invited to join validator-specific communication channels. The successful completion of the onboarding will be announced here, and the relevant documentation subsequently updated.

3 Likes
61

The onboarding of the 21 selected validators is now complete.

Selected applicants have been added to the Validator Registry on-chain, and are now receiving equal shares of newly staked Luna. Despite concerns surrounding the sunsetting of the Terra Delegation Program, all selected validators were in the active set at the time of the onboarding.

As part of the Stake Flattening Procedure, 180k Luna have been redelegated to each entrant validator, for a total of 3.78M Luna. This is in line with the current SFP requirements, which limit the total amount of Luna to be redelegated per increment at 4M Luna. We will conduct an internal evaluation of the SFP, which could lead to a proposal with parametric changes to ensure a smooth and efficient flattening of stake across the expanded validator set.

The whitelisting transaction can be found here, and the results of the onboarding there.

2 Likes