Updates on Test Data

Announcements
903

This was a concern in the past already: Do full nodes get vetted? - #5 by jammerdan

1 Like
904

With SLC hammering nodesā€¦ is it interfering with US1 repairs? 36 is No Bueno for data customers are paying forā€¦

2024-06-06 21_25_17-US1_repair

4 Likes
905

Our devops team is already aware of this issue, but thanks for pointing it out.

5 Likes
906

According to the Community Forum Code of Conduct Code of Conduct this post is your first offense warning under the Hateful Conduct and Harassment clause due to your choice of language.

Please take care to communicate in an appropriate manner when expressing your thoughts in the Forum

7 Likes
907

Hence why I had to infer why he posted it as a solution to the large GC run I was seeing.

908

SNOs who have 50+ nodes and 100+ TB of data can already afford to have normal router, it is not so expensive at all

3 Likes
909

You donā€™t need fancy router.
My asus ac3100 was able to handle the 500mb/s without any problem.
Will be upgrading to 1gb or 2gb (will need new router) down the line if this becomes the new normal.

1 Like
911

Thatā€™s the key here. Itā€™s much easier to replace routers with more ram, than to move to entirely separate infrastructure and essentially build parallel internet.

And I donā€™t think that memory requirements are drastically different ā€“ stateful firewall still needs to keep track of pretty much the same amount of data.

IPv6 indeed should have been designed with built-in interoperation with ipv4. It wasnā€™t, and thatā€™s why ipv4 stays.

3 Likes
912

I use Pfsense. Itā€™s free and can handle everything with ease.
So good gear donā€™t have to be expensive

5 Likes
913

I donā€™t get the router argument. For stupid lock in reasons I am using my consumer ISP router in conjunction with a consumer Asus router and both held up just fine dealing with these tests, despite getting the traffic from 6 /24ā€™s. This is not exactly the most efficient setup and yet, Iā€™ve seen no issues.

4 Likes
914

NAT is not a issue at all.

2 Likes
915

This!

just buy a fast 4-8 cores with 3 ghz + and you pull over 10-20gigs pure l3 with ease. Lets drop the network discussion. Its not an issue, crazy load will however shorten the lifespan and aloy of deletion of small files combined with high ingress of files. Will be hard on spinning disk so that something everyone needs to take into account.

2 Likes
916

Hi,

So I am a 3 week old SNO, running enterprise gear that runs 24/7, so allocated space to the project. Like others, have seen a huge amount of traffic as a result of the testing.

As I said, pretty new to this, but just wanted to understand what this report is telling me in light of these changes. I was seeing small increasing storage each day, before ingress picked up on the 4th July.

But can someone help me understand the rational on these numbers, maybe I am missing something here. So for all the 3.4TB of ingress, 1.7TB stored, that nearly all of it went to garbage after being stored for a short time - for which no payment is due? Not paid for ingress, and not paid on short lived storage either. Have I interpretted that correctly?

REPORTED BY     TYPE      METRIC                PRICE                     DISK  BANDWIDTH        PAYOUT
Node            Ingress   Upload                -not paid-                        3.45 TB
Node            Ingress   Upload Repair         -not paid-                        0.00  B
Node            Egress    Download              $  2.00 / TB (avg)               10.51 GB       $  0.02
Node            Egress    Download Repair       $  2.00 / TB (avg)              191.85 MB       $  0.00
Node            Egress    Download Audit        $  2.00 / TB (avg)               76.03 KB       $  0.00
Node            Storage   Disk Current Total    -not paid-             1.69 TB
Node            Storage              ā”œ Blobs    -not paid-             1.65 TB
Node            Storage              ā”” Trash  ā” -not paid-            36.96 GB
Node+Sat. Calc. Storage   Uncollected Garbage ā”¤ -not paid-             1.23 TB
Node+Sat. Calc. Storage   Total Unpaid Data <ā”€ā”˜ -not paid-             1.27 TB
Satellite       Storage   Disk Last Report      -not paid-           423.63 GB
Satellite       Storage   Disk Average So Far   -not paid-           203.00 GB
Satellite       Storage   Disk Usage Month      $  1.49 / TBm (avg)   33.41 GBm                 $  0.05

Thanks CC

917

All data, either normal or TTL, is paid proportionally. If you store something for a day, you get 1/30 of the monthā€™s payout for it.

6 Likes
918

Not exactly. if you have a multinodes system that works great under 1gbit average before testsā€¦ than you push it to 5/10gbit limit. If you had ā€œconsumerā€ router you reached 50.000maximum connection or other limits (qhora and similar router). So you forced to upgrade to 1/2k usd router family or just build a DIY pfsenseā€¦ btw. you forced to fix a problem.
Iā€™m happy with thisā€¦ I was forced to evolve myself and my network knownledge :zipper_mouth_face: :rofl:

2 Likes
919

Clipboard02

This is fine? This shows on all nodes.

920

But who does 10gbit on a consumer router, is there any consumer routers for 10gig besides maybe unifi?

921

Same here on all my 60 nodes.

922

Its fine, just means that a satellite (in this case it seems to be US1) hasnt been able to push avg usage stats to nodes for those days. Actual payout should be unaffected.

2 Likes
923

Like this?

2 Likes