Pre-Dev-Update Thread! Yay! :D

What are the odds for a T5 today?
50/50? 1/4? 1/10? Place your bets please.

50/50 ! (10 character)

3 Likes

20% but I hope I’m wrong

Only without Anti Entropy I think, and since I don’t think there’s much point in that I don’t expect T5 for at least another few days.

3 Likes

More code added → more things can fail.
When many things fails simultaneously, testing becomes harder.

(just describing why it may have sense)

2 Likes

Yes, and an example which might be worthwhile testing is DBCs if they’re ready. I’m still thinking next week though.

4 Likes

Anti Entropy being a bit pernickety?

Could be too much of a tough nut to crack this week :weary::weary:

Hoping for the best :muscle:

2 Likes

Update from MaidSafe HQ

We’re tackling what we think are the last couple of bugs before testnet v5 is considered ready to release.

The main stubborn bug is that something in our Anti Entropy implementation seems to be causing high CPU usage. We’ve got the guys digging to find out why & resolve, and hopeful that will be done in time to release tomorrow.

Thanks everyone

48 Likes

Never happier to be proven wrong and this is becoming a habit. Go MaidSafe!

23 Likes

From running the latest sn_node on my local Baby-Fleming, I can confirm CPU usage is bit severe right now.

8 Likes

This comment reminded me of this cartoon

13 Likes

@StephenC @bochaco When T5 does come - and there is NO pressure from me to make that happen tomorrow- just want to make that clear :slight_smile:

Would it be a reasonably safe bet to say that the network config file would be found at
https://sn-node.s3.eu-west-2.amazonaws.com/config/node_connection_info.config ?

And of course the contents of this URL at release time will not necessarily be the same as they are now - so there is no point in peeking yet.

Asking for a friend who would like to get ahead of the game with a screencast

8 Likes

I think @StephenC was aiming at always publishing testnets bootstrap addresses at that same S3 URL

7 Likes

P p. Nm l lol L ml

Edit: seems I was pocket posting. Sorry. Haha.

1 Like

I upgraded to the latest node v0.41.1 and it seems much faster

willie@leonov:~/projects/maidsafe/GETs$ time safe files put /tmp/std-file-resources/testfiles/5MB.zip 
FilesContainer created at: "safe://hyryyrbmwexwsdt53durbxrwrkttixpauy1pfkrpu7qpsshnzji4jk3znjcnra"
+  /tmp/std-file-resources/testfiles/5MB.zip  safe://hyfenrym6m19mkdjperroq88n14mkpbemyatwsxehwdrc6rhnqpcemeud3y 

real	0m37.677s
user	0m8.591s
sys	0m0.475s
6 Likes

Also less CPU intensive, I presume?

How it relates to non-merged PRs like #1479 ?

1 Like

We have a new sn_node now at v0.41.2

Again seems a lot faster – this is on an 8-core i7 16Gb RAM Ubuntu 20.04

willie@leonov:~/projects/maidsafe/GETs$ time safe files put /tmp/std-file-resources/testfiles/10MB.zip 
FilesContainer created at: "safe://hyryyrbk9cucg3xrdkmaotojxq5afsj4ad9qcrhs1tk4pcbkffe47ync5dcnra"
+  /tmp/std-file-resources/testfiles/10MB.zip  safe://hyfenrykcrjmkk78txsiep9zzh9a6ah5jgccjuo8g9yeicjkimwsp8hc31o 

real	0m54.403s
user	0m9.143s
sys	0m0.824s
willie@leonov:~/projects/maidsafe/GETs$ time safe files put /tmp/std-file-resources/testfiles/20MB.zip 
FilesContainer created at: "safe://hyryyrbpwsa4trmo73jwf6nk493mchedfarta7ryrdxkkef1611yrmti4kynra"
+  /tmp/std-file-resources/testfiles/20MB.zip  safe://hyfenryxon8oeymwyeeau8cgt3nj5u1g9psuzqojthz5n4pmz95jxxg93uh 

real	2m0.788s
user	0m24.419s
sys	0m1.759s
willie@leonov:~/projects/maidsafe/GETs$ ^C
willie@leonov:~/projects/maidsafe/GETs$ time safe files put /tmp/std-file-resources/testfiles/5MB.zip 
FilesContainer created at: "safe://hyryyrbe3411cfnrqds7jdfcczw6j9qihneowdwxokrsxhb83hkz5jm3u7anra"
+  /tmp/std-file-resources/testfiles/5MB.zip  safe://hyfenrym6m19mkdjperroq88n14mkpbemyatwsxehwdrc6rhnqpcemeud3y 

real	0m41.759s
user	0m9.662s
sys	0m0.446s

It seems faster and while the CPUs were all at ~100% for the duration, RAM usage was low.
Need more testing with std files to get a feel for this across different architectures.

2 Likes

There’s gonna be another soon :wink: Also incoming sn_client changes (just small)

7 Likes

I’ll hold off on repeating that test on an older 4-core i5 then…

BTW just WHY are you doing this on an election day? I need to be in several places at once. If it wasn’t raining and turnout low in the afternoon I’d be out doing work for the Boss :slight_smile:

2 Likes