Dencun Interop Testing #40

Transcript

cool I guess we can start um yeah and maybe anyone from the Cent teams want to give an update uh where you guys are your releases um are they already done is it planned this week Tech already released and we may release another another version before mainnet not sure but current release is already ready well star we have already cut a RC and we'll release it either today or tomorrow and we'll basically might have some followup Improvement releases but

not really related to D okay good um would you guys want us to wait for your followup release for the main night Shadow for or can you go ahead with the RC release you can go ahead with the RC release I can share you the RC version awesome that sounds good do we have any other Cent team here anyone else um want to give us an update on releases yes so NE mind already released Friday uh and uh on Friday spamming we found an issue with not keeping capacity of blob pool under heavy spamming I'm

working on it right now and it's likely that after fixing it we will release one more version but it's not critical so uh we like release on Friday and there will be no problem if people will go with it the hard for could you give us an idea on how much you had to spend to qualify spamming like how much e was spent to cause the issue when I'm using my spammer I'm organizing like 500 if from testet and using into to spam but I think if I would collect it from Accounts at the end it will

still be 300 or 400 left so so the real cost would be lower and most of these transactions it's possible to send transactions which will not be uh included really so if you send like a lot of transactions with Max blob gas equal one it will probably never be included but you can spam with it hardly oh go yeah that makes sense thank you and I think Barnabas was also looking into something with get on the weekend right yeah that was some performance

issues also uh regarding blob transactions their block building uh took over 2 seconds and some of the blocks were not proposed because of that there's a fix now and it's been rolled out and it seems to be working but uh we've been spamming with the BL fuzzer and uh it doesn't seem to be able to recreate the same issue so it would be nice if we could have another set of large blob spamming experience yeah I repeat the Friday experiment for sure like for now uh I'm trying to just fix our issue to to test

it on on already fixed Version Y sounds good awesome so that probably mean there also another get release um over the next week I guess uh I would assum so yeah oh yeah I already see it Ted to the next Milestone okay good um I also see there's a bzu release that's been done web3 signer is done as well um yeah and the other clients are stilling I guess do we have anyone else from the other CL teams I think the CL teams

mainly said they'd be making releases this week yeah I think ter mentioned that prison will be ready in Thursday okay but Cas just joined so maybe he he has an update for us too just Forex case you were asking about releases I see yeah um we are just getting started with soaking um so that you know actually I don't have a solid answer for you um we're we're testing a few things out right now but that is our that was our original goal so

hopefully that sounds good okay was there anything else that was found this week that we want to talk about uh babus you want to maybe give an update on Deet 12 I mean there's not much happening other than the blob spamming and the g issue that we found but uh we seem to have very high uh participation right right now on D 12 so I think most of the clients are pretty okay yeah that sounds good um yeah I

guess that probably means means we're still waiting for the client releases hopefully that happens this week and if that does then we're looking at Main net Shadow for this week otherwise we'd be looking at me at Shadow F next week right and I think Tim wants to wait at least till or we agreed that we wait at least till the shadow for is done before the block post goes out regarding the shadow Fork it looks like all of the clients have uh synced up except Aragon none of the Argon notes have suned up and uh one rest note but all other pairs seems to be fine yeah that's good um they think we

have similar answer for the girly sync test as well cuz that's beyond the uh blob expiry window and I think ericon might still another day that's like 10ish per so I still give it a d and see um what happens there um I see we also have uh people from SAR to stret could we maybe get an update on if you guys have started testing blobs is everything going okay I saw there was some governance proposals on optimism for example yeah I can give an update on that um so we we've made a proposal of governance

toet um I think three days after dun um 18th of March um so far so good no objections um but you know don't know exactly how that's gonna turn out um on T I believe it's Tuesday we'll be upgrading sepolia to that particular op stack release this is called the the eone release for op stack um it's already running on op gly um and it's we've done a couple devet hard forks with it as well so we believe it's ready to go um you know there may need to be a couple last minute tweaks but we don't expect any big changes in it so overall in pretty good shape so this is B for both base mainnet as well as op mainnet

that would be the March 18th date awesome thank you so you were saying something us um I guess one followup question I had was um I I have heard from other people asking as well that uh tradition RPC providers have only provided El rpcs um do you guys have the same problem and if so where are you getting the blob data from IE where's your do you do you know if RPC providers are having issues offering the beacon API uh we haven't received a lot of

feedback on that to be honest um you know we run our own RPC providers our own RPC nodes so um not entirely sure but we also work with quick node and they haven't really expressed any concern about it but um it's possible I'm not in the loop there okay it might also just be one of those things where they just have to spin it up yeah some and um are you guys already looking at Hive Cancun to check for field tests for example I think there

still are some um engine Cancun Fai test on another mind for example also death um so from nether M side those H those hes are non issue not related to K and on the solution is on PR uh I think this one which you link are a little bit outdated it's not the current result of five tests the image the image on hive is updated or Hive is

updated so I'm not sure we have our workflow and uh we have totally different results I think on hi cancon uh there was image nether mind canun used from K branch and right now we are working master and it still executes very old test and I think there was there was one page which execute um the new test but let me try to find I can change the source for sure um we can ask Mario oh oh I have I have this page I think yeah I think I should

also have access to make deployment change St so it should be easy it should be running the the right tests I think just there's an issue with running the tests on hive Cancun when I run locally there's maybe five I think there was five test failures for NE mind but nothing major I think there were all error code changes oh yeah yeah error code changes are Fe on Branch it's not Mer but it's totally nothing that we should worry about and as you can see on this High test two page ah but this is very no no it's it's uh fresh executions we are passing

all of the tests hi to hi Test 2 is better than hi canun uh am I correct sper yeah they run on different instances so definitely isely yeah there there's something weird that's going on that uh has been a struggle to find out why why the some of the tests are failing maybe it's running a running the parallelism parallelism too high um I think when we overhaul Hive and the hive c will aim to fix that but it's not really been a huge uh

priority and yeah sounds good um it also looks like they're running different commits so it could just be that one of the commits CH something uh I think we're still on oh no we're not on a for okay um you can dig into that later but I guess if people are running it locally in in their own CIS that's the best view of what's going on okay is there anything any other

topic people wanted to bring up today I don't think there's anything in the agenda besides releases for today okay I guess if there's nothing else then you can end the call have a nice day everyone thanks cool bye thanks guys