Updates on Test Data

Announcements
149

Just for safety I better post it in this thread as well:

We will do some spot checks regarding the bandwidth our storage nodes have. We will run the new remote benchmark tool against some random selected nodes for a short duration. We are going to use big piece sizes to compensate for the fact that the performance improvements are not rolled out yet. We hope we will still get an accurate reading. The data will be unpaid because we upload the pieces to the storage nodes without commiting the result to the satellite. Garbage collection is going to clean it up later. We will keep the upload burst short to keep the amount of unpaid data as low as possible.

11 Likes
150

If youā€™d like more nodes to be brought online in specific locationsā€¦ will that be combined with your SLC-capacity-reservation-with-TTL system?

Making an announcement for more nodes somewhere is a relatively weak signal to SNOsā€¦ but focusing capacity-reservation at existing nodes in your required geo would be a strong one. If you want a extra nodes brought online in a geo: start filling the disks of current nodes with your SLC reservation data :money_mouth_face:

2 Likes
151

Roxor, Yes we have considered this approach and will likely implement it along with an announcement at some point in the future.

2 Likes
153

We just got our first performance test 5-6 hours ago? Looks lovely in the traffic graphs: hope you saw the numbers you were looking for!

154

Something overwhelmed my Pi-hole a few hours ago. I guess it was the storj testā€¦

I had to change the ā€˜dns-forward-maxā€™ setting to a higher value, default was 150.

155

Storagenode rollout should be finished shortly. Next round of benchmark tests incoming. Prepare for impact. Everyone hold tight :smiley:

8 Likes
156

Meanwhile my Windows node.

Compared to Linux node

1 Like
157

I believe we managed to reduce the gap between windows and linux performance. Lets see how your windows node performance now :slight_smile:

2 Likes
158

Hm, is this good or bad news? :thinking:

1 Like
159

2 of my newest nodes never receive a single vetting attempt from SLC. The rest of the satellites have been fully vetted for months on these nodes. Iā€™m assuming because of this situation (no vetting) they wonā€™t receive any of this test data. Pls confirm.

160

Well, in the lat 30 minutes all my nodes have drastically reduced network traffic.
So either Iā€™ve missed it or the tests have been happening and are now completed :man_shrugging:t2:

2 Likes
161

I donā€™t think theyā€™ve started the tests: soā€¦

hold_on
hold_on825Ɨ530 55 KB

1 Like
162

Then why has this happened?
If it had been just on one machine Iā€™d ignore it but these machines are spread across 3 different countriesā€¦

image
image1920Ɨ1200 156 KB

1 Like
163

It looks like they chopped all synthetic traffic (or broke a satellite?). Even large SNOs had traffic go to near-zero.

1 Like
164

These are some beautiful colors. Unfortunalty I am bad in mind reading. What do you mean with ā€œthisā€?

1 Like
165

You wouldnā€™t want to read my mind, there is not much thereā€¦

I was alluding to my earlier post about severely reduced network traffic and node activity (which you can see by the reduction in all the pretty colours on Disk IO, IOWait, Network, etc.) :wink:

166

Perhaps itā€™s the calm before the stormā€¦ :wink:

Bring it on, guys! Iā€™m curious to see how far you can push these things. :muscle:t3::smile:

1 Like
167

It looks like traffic is climbing back again (to already half of what it was an hour ago?). Maybe some Storj Intern rebooted the wrong system :wink:

168

Maybe wait a few days until all BF are processed; the Trash is still cleaning. My Synos are boiling like never before.

2 Likes
169

The network has to handle it even while garbage collection is running. So perfect timing to test it out.

5 Likes