And regular egress is very bad at the moment.
2 Likes
In fact I have days where I have overall more repair egress than regular egress. 
Where are the egress use cases that Storj is going after?
3 Likes
suddenly traffic go south and south and south…
1 Like
What got me wondering is, why is repair inbound and outbound constantly this high? Shouldn´t it only kick in, if there are not enough pieces?
Well it would be a problem if repair kicks in when there are ‘not enough pieces’. That means that there would be data loss. But you are correct in what you probably mean: when the amount of pieces falls under a certain threshold, repair kicks in the bump up the pieces to a certain amount again.
I also noticed repair being higher since the start of this month. That would indicate to me that the churn is higher since the starting of the new payout?
1 Like
Repair egress has been 2-4x higher than regular egress each day since the start of this month.
(on my nodes)
My guess is that there are enough people exiting that we’re seeing the repair go up. I assume once it all settles we’ll go back to normal.
Almost half the ingress (from ~40Gb to ~25Gb), egress repairs went to practically zero.
1 Like
Same. Literally when I wrote my comment, the repair went down. Who knows if the data is accurate at all.
1 Like
Same here, repair is at 0 and it almost never happened before on my oldest node.
(post deleted by author)
EU north and saltlake are only test sats they dont really matter.
1 Like
On the other thread which discusses this topic I see you yourself have also posted some comments, and our devs have responded to concerns already there 2 days ago. No "Disk Space Used This Month" since yesterday - #14 by kaloyan
So please abstain from spreading FUD. Devs are actively addressing some issues with performance which as a side effect resulted in the weird looking statistics for 2 test satellites.
Note that this is why test satellites exist, to test changes before applying them to production satellites, so as not to affect customers. If something does not work as expected, then the devs can work on fixing it on the test satellites without negatively impacting customers on production satellites.
2 Likes
(post deleted by author)
- Devs have indicated that they are working on some performance issues, their testing resulted in the displays you described as concerning on 2 test satellites.
- you posted on the same thread where the Dev posted their comment, therefore, you are already aware that they did respond, while claiming that there is too much silence from the company.
- Please indicate if you prefer devs work on testing new code or chat here with you instead. One comment indicating they are addressing an issue should be sufficient.
1 Like
(post deleted by author)
I think it might be worth making a pinned post addressing what’s already been said.
Currently it’s buried in a thread that doesn’t exactly have a title that makes it easy to find answers.
I didn’t find it until you pointed me to it after I made a post asking questions about it.
“Updates on storjstats” or something along those lines.
First, Heunland is an employee of Storj Labs, she isn’t White Knighting anything. She is correct, that if you posted something, and someone responded to you to answer your question, it would make sense that you would see the response and not just bring up the issue again in a different topic. Otherwise, if you aren’t interested in answers, don’t ask questions. That said, it is Ok to ask about potential issues you are seeing, we welcome this kind of feedback as we’re not always aware of everything everyone is witnessing.
Second, it is possible something is being worked on right now which may be causing some hiccups. We don’t get briefed on everything the various teams are working on unless they feel it warrants letting everyone know. Partially because we don’t want anyone exploiting the issues we are working on. And also because Storj is a small company and resources are better spent working on issues and not crafting messages to explain about hiccups. I think everyone would agree that if there happens to be a problem, that the team should be working on fixing it first, and then can tell us what happened later if necessary.
At this time though, I have not heard of a problem happening. There are many different teams at Storj working on different things. Everyone is remote and in a lot of adhoc meetings. So, it’s not always easy to know what everyone has going on until we get a break down in a meeting or subsequent company communication. If something like that comes out, one of the team members will share it. But most of the time you can simply check the release notes to see which issues were addressed.
10 Likes
From my experience, the traffic seems to be somewhat resolved with version v1.78.3.
Yeah, development is not and easy thing specially with the complexity of Storj, hiccups will happen more often than not, I think anyone that has some experience understands that.
Fixing the problem is more important than rolling heads, otherwise the problem will never be fixed.
Just keep working, ignore toxic comments and be transparent to the best of your ability like you have been.
2 Likes
