What is “InvalidArgument desc” ? i got those all the time now(solved)

Ruskiem · October 6, 2019, 3:59pm

Hi. What is “InvalidArgument desc” ? i got those all the time now, after satelite being unreachable what some people reported here (118UWp satellite seems to be down), now im online but with plenty of these:
for example:
“2019-10-06T22:41:53.888Z ERROR server gRPC stream error response {“error”: “piecestore: rpc error: code = InvalidArgument desc = order created too long ago: 2019-10-06 13:41:51.771361948 +0000 UTC”, “errorVerbose”: “piecestore: rpc error: code = InvalidArgument desc = order created too long ago: 2019-10-06 13:41:51.771361948 +0000 UTC\n[tstorj.io/storj/storagenode/piecestore.(*Endpoint](http://tstorj.io/storj/storagenode/piecestore.(*Endpoint)).doDownload:419\n[tstorj.io/storj/storagenode/piecestore.(*Endpoint](http://tstorj.io/storj/storagenode/piecestore.(*Endpoint)).Download:365\n[tstorj.io/storj/pkg/pb._Piecestore_Download_Handler:1096\n\tstorj.io/storj/pkg/server.(*Server ](http://tstorj.io/storj/pkg/pb._Piecestore_Download_Handler:1096\n\tstorj.io/storj/pkg/server.(*Server)).logOnErrorStreamInterceptor:23\n[tgoogle.golang.org/grpc.(*Server](http://tgoogle.golang.org/grpc.(*Server)).processStreamingRPC:1127\n[tgoogle.golang.org/grpc.(*Server](http://tgoogle.golang.org/grpc.(*Server)).handleStream:1178\n[tgoogle.golang.org/grpc.(*Server](http://tgoogle.golang.org/grpc.(*Server)).serveStreams.func1.1:696”}”

Its win 10 pro.
( i had no problems at win 10 pro up this time, no docker problems, i cant complain)

Alexey · October 6, 2019, 6:01pm

Perhaps your node were offline for a long time and the order is missed its expiration date.
How long it was offline?

Ruskiem · October 6, 2019, 6:08pm

like 2 days? i was trying to change hdd, and got errors, so it took me 1 or 2 days to instal windows and docker again, just later to find at forum that wasnt my fault coz the satalite was down, so i wasted time for long windows instalation but i least i have it on smaller hdd

Alexey · October 6, 2019, 6:10pm

Yes. Because your node should not be offline for more than a 5 hours total in month.
Please, read this blog:

Ruskiem · October 6, 2019, 6:13pm

Heeey im not new? im here from 2 months!

The blog says about "Containment mode ", nice, cant wait for it.
I never got this “invalid argument desc” and been offline for longer even. My node is online, will he recover by itself? no traffic makes me saad;( or i got to do something? he keep getting those errors.

stefanbenten · October 6, 2019, 6:38pm

Orders expire after 24 hours if i am not mistaken.
Meaning that all orders that you stored from the time before are invalid.

To ensure that i am right, lets wait for an confirmation of @littleskunk here.

Alexey · October 6, 2019, 6:41pm

Containment mode is implemented a long time ago. You can read a documentation here:

github.com

storj/storj/blob/eb5413ae5e9b5675007f7c6ee0acec6f68e87188/docs/blueprints/audit-containment.md

# Auditing and Node Containment

## Abstract

This design doc outlines how we will implement "containment mode," in which the auditing service monitors a node that has received an audit request, but refuses to send the requested data.
Currently in the codebase, nodes that do this will be marked as having audit failures.
However, we don't want to immediately mark an audit failure if this specific case arises, because factors such as unintentional misconfiguration by the SNO, or a busy node fulfilling requests should not necessarily warrant the penalty of an audit failure mark.
Instead, we want to "contain" these nodes by retrying the original audit until they either eventually respond and pass the audit, or refuse to respond a certain number of times and ultimately receive an audit failure mark.

## Main Objectives

1. Nodes won’t be able to “escape” audits for specific data by refusing to send requested data or going offline right after receiving a request from a Satellite.
See _Identifying Nodes That Need to Be Contained_ subsection below for how we consider "refusing."
2. The audit service should attempt to verify the nodes' originally requested data in addition to continuing normal audits for that node.
3. ContainmentDB holds metadata required to verify contained nodes' originally requested data.

## Background

The whitepaper section 4.13 talks about containment mode as follows:

This file has been truncated. show original

Ruskiem · October 6, 2019, 6:51pm

wow, im sorry, havent seen anywhere how to actually use that Containment mode, on winndows. Actually i was trying to get online for that 2 days, a hdd switch was under 5h for sure, but i got false belive that i did something wrong, later to see it was satelite down, making me thinkg i F UP something. Now i want to recover that beast faaast, my reputatyion was unblemished ;D

Alexey · October 6, 2019, 6:53pm

You can’t force it. It’s automatic, if your node is failed audit, it placed to that mode automatically to get a second and third chances to respond on audit

anon27637763 · October 6, 2019, 7:09pm

I’d like to point out that the blog post you linked to specifically indicates that “Containment” mode is not used for SNs that go offline. Those nodes are marked “offline”…

So, I’m not sure containment mode applies in this particular case… but maybe the policy has changed since the blog post.

Alexey · October 6, 2019, 7:13pm

Many people have had an assumption that this mode could be used as a maintenance mode, but it’s not true.
This is technical term for placing node to some kind of isolation from the traffic until the status of audited pieces will be resolved. Do the node have this piece or not. Should the satellite treat this piece as lost or not. It’s designed to solve two problems: do not trigger the repair because the node has had temporary difficulties to respond in time (was busy for example with uploading a piece to the customer) and protect the node from losing audit because of load.

anon27637763 · October 6, 2019, 7:14pm

From the github document:

However, we’re departing from the whitepaper because offline nodes will not be moved to containment mode. They will just be marked as offline. Instead, containment mode is specifically for nodes that initially respond to the audit service’s dial but then don’t send the requested erasure share.

littleskunk · October 6, 2019, 7:17pm

Without containment mode you might get an audit failure for restarting your node. The containment mode will make sure you get a second and a third chance to avoid the audit penalty.

It has nothing to do with offline penalty!

Ruskiem · October 6, 2019, 7:19pm

Guys guys, great! but i have 0 audits (audits_satellites.ps1 shows) from 19h, should be some at least! is something broooken?;>

anon27637763 · October 6, 2019, 7:19pm

OK…

But that statement departs from the github post that was linked to earlier…

So, I suggest that the github document be updated to reflect the current behavior.

littleskunk · October 6, 2019, 7:20pm

Orders are valid for 7 days. You can submit the order even days later.
The error message you see has a different reason. The uplink has only 1 hour to finish the upload and send the pointer to the satellite. This is needed to avoid some special attacks an uplink could execute otherwise.

Alexey · October 6, 2019, 7:21pm

If you do not have data from that satellite, what it should audit?

littleskunk · October 6, 2019, 7:22pm

The statement on github is correct. Especially the last sentence.

Ruskiem · October 6, 2019, 7:22pm

i have plenty of data, this month was good, like 160GBs, i have fully vested node, ver 500 audits from every satelite (3 of them) but im online 19h and not even a one, so i worry ;>

Alexey · October 6, 2019, 7:23pm

I tried to upload a huge file, if it could take longer than a hour what would happen?