Node offline for 20 hours

Hi
Yesterday without any apparent explanation, electricity went out of part of the city. With everything and that I have a UPS, it was not enough its 2 hours of battery to keep the node on. It was almost 20 hours that my node was offline.
It is a new node, it has approximately 20 days active.

Now the electricity is back, what do you recommend I do?
1.- Continue with the node with everything and the almost 20 hours it was offline?
2.- Create a new node and start again?

Comments in this regard are appreciated.

it should be disqualified by now, since the max offline time is 5h but…
I recently had a problem and the node was 6h online. Adding all I´ve been offline for 8-9hours last month, but I wasn´t DQ, so… I would wait and see if you are or not disqualified.

If DQ, you should start again.

DQ because of downtime is not implemented yet as far as I know, only for too many failed audits. So you should be fine to just restart.

2 Likes

you’ll be fine even with 20 hours downtime. No downtime DQ active yet.

Otherwise the dashboard would tell you and you would get an email if you got DQed.

Is this DQ following extended offline going to come back? If it is, I might as well shut everything down now, because turning the power off for a whole day happens here. Last time it was 9-5 on 3 days in one week for line maintenance. They also cut the fibre cable, because it was in the way and they didn’t bother to tell the ISP.

It is going to come back: Blueprint: tracking downtime with audits
Not sure when exactly. First it will be implemented for testing without real DQ and once that works, they’ll switch it on.

Are you expecting outages like that more often in the future?

Yes, I do. It’s just a matter of when.

Looks like 3rd world countries need not apply.

You should stick around anyway. There have been quite a few indications that this down time disqualification won’t be nearly as strict as it used to be and there will be a time of suspension before disqualification to give your node time to recover. I can’t promise you it will be enough for your situation, but I definitely think it’s worth a try.

2 Likes

Thanks, I probably will. Nothing to lose, anyway: my rigs were bought for Burst crypto mining, which needs lots of HD space and that doesn’t even make me the marginal cost of electricity, above what my solar array generates. I even have a small battery bank and UPS, but then last week the RCD tripped after midnight, taking out the entire house, incl. fridge and freezer - two days in a row.

Hi,

Would it be possible to create a thread to give us some insights about the time Storj Labs will implement and active the Downtime DQ?
I plan to move in another countries in the next few months and it would help me to chose the best option.

I suppose Storj Labs teams still don’t know when they will implement it but it would be nive to include it in the following thread, for example: Known issues we are working on.
This way, Storj Labs team could tell us when it will be implemented once it will be decided.

Thank you!

SNO suspension due to downtime likely won’t be released for another few months. In addition to wrapping up the implementation details, we’re concentrating on getting the best performance and durability that we can and we don’t want to suspend nodes for downtime until we are sure it won’t unnecessarily penalize nodes. I’ve also made a note of this in the “known issues” post.

5 Likes

Appreciated! :slight_smile:

You guys should really consider who the Storj project is aimed at SNO wise. If you want normal private persons (i.e. with a dayjob) to participate, you can’t punish them for, say, being asleep.

When one of my nodes goes down for whatever reason 1h after I fell asleep, I won’t likely notice until I check my Telegram notifications at some point after getting up. Even if I’d see such a notification at night, I will probably not get out of bed just to check the issue out. This being said, everyday I go to sleep at night, I risk permanent DQ to my nodes (assuming downtime based DQ was in effect).

I personally like the fact the Storj currently can be participated in by both datacenter owners as well as someones mom’s laptop with an external HDD. Please don’t shut out us smaller SNOs :wink:

+hardware failures tend to take way longer than 5h to repair, since a lot of the times you are gonna have to order new hardware in those cases…

/edit: and no, I don’t run a node on my mom’s laptop.

1 Like