New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Comments
Guys, the end is nigh, right? Guys??
I think, all LUX node..
We're heading towards the 6 hour mark on the slices and no sign of Fran. Looking forward to the RFO
A full day of downtime of my vm slice with @francisco ... A bad first experience for me
So not only LU shared is down but also LU slices?
Yes, had 4 Lux slices in different nodes..
But it seems already up..
I'm still debugging what happened, but i'm assuming it's some bug in the MX204 firmware?
I recently cut LU over to a new router and it just...stopped forwarding TCP traffic. ICMP was fine, but TCP just fell flat on its face.
Since ICMP didn't fail it wasn't setting off alerts enough.
Sorry, entirely on me for not having even more alerts.
Francisco
@Francisco
Two months free for all clients
Deadpool.
>
Never heard of such a bug and we run a ton of these? What code are you running in it? What did JTAC say? For own professional betterment will be interested to know a PR number as well so if there’s an issue with this it can be fixed before it happens, if that really is a code issue, that is?
You can ticket and Karen will kick you a week of SLA.
Francisco
I have a VPS in Lux, it's up right now.
By the time I woke up things were out for a long time so I simply sent it for a reboot to get things going again. I've not opened a case on it, but will look into doing so.
We're running "17.4R2.4" on it. I want to upgrade to 18.x for SFLOW and such, but haven't done it yet.
Francisco
Do you happen to have long firewall chains, say in terms of prefixes or rules ? 204 has a significantly more limitations in that respect, compared with say MPC3/7 on 480. It’s possible that part of the chain wasn’t implemented resulting in what ended up to be a partial ACL that resulted in not allowing some traffic in rules further down the chain. If so, I have a PR for this somewhere, I know our usual ACL caused it some issues, simplifying the firewall and reducing the number of prefixes in the terms fixes it.
Our management ACL is only like 6 TERMs long, very basic port/srcip/proto combos, nothing fancy at all.
I'm keeping an eye on it, but it was really odd. Packets over 300 MTU would fail too.
Francisco
Something something MTU set to 500 instead of 1500 ...?
EDIT2:
aha, missed that - nevermind ...
Looks like it's back online now. Shouldn't be a problem anymore.
You wont get your Provider tag if you keep spamming with useless comments.
I was curious when he started at Frantech. He seems to know more then the rest of us, so I figured he worked there.
🐐
Fran doesn't need any help, he can do it all. He's a pony man!!!
Offline again?
Edit: tagging @Francisco
yes, my vps is off
Is it VPS only or webhosting?
Oh for fucks sake
https://status.foxo.me/783135858
Edit: this time though not even ICMP packets go through.
I only have VPS
Francisco go to sleep
All hell break loose
now, only Slices down for 1 hour n counting..
I believe it is VPS only. Lu4shared is up
I feel bad for @Francisco! Hoping he and his crew finds a solution soon. I know the feel and pressure in a situation like this, especially when you think hell was resolved...
Until then, I will be waiting until the slices are up and we get stable service again!