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
@Nick_A
We're running the Solus panel from yourselves, can you tell me the point of data loss? So I can manually recreate entries into Solus for clients etc.
I believe it was around 6 AM EDT.
The 2 nodes I am on had 231s of downtime, pinged every 5 seconds.
The real downtime was minimal... Pingdom is not a good factor of judging real uptime/downtime. It was really ~4 minutes.
Depends on which node you're on.
I assume we're talking about the network dropping out and nLayer not kicking over? How does that depend which node you're on?
He means some VMs were on a node that died and experience extended downtime while they are moved to a new home, therefore downtime depends on node too, tho it is unrelated to the network issue.
Right. I checked their Twitter, wasn't aware of the failed array earlier today. I was speaking about the network drop, few people made it seem it was longer than it really was.
All I can say is that all of my RamNode VPS were having network issues for five minutes and then were down for thirteen minutes. I don't know what happened with the node(s) your VPS is/are on. So that's eighteen minutes in total. Not making anything seem longer than it really was, maybe you just didn't have the same issue; not sure why you're trying to say that my VPS was inaccessible for four minutes when I know it was offline for more than that.
This is based on both being SSH'd into the VPS at the time (so during the first five minutes, freezes of thirty seconds or so, and then it went down for thirteen), looking at connections disconnecting from the running services at the same time as the SSH problem (means it wasn't just me), and also uptime reports from one of the VPS showing that it was unable to ping several (every) external locations, with on and off for the first five minutes and then continued through the thirteen minutes.
(Not saying that this is bad or anything, this is pretty rare for RamNode. Just saying that it was >eighteen minutes, not four/thirteen; and irritating that "few people" are claiming this is incorrect without full information.)
Irrelevant, eighteen minutes wasn't based on Pingdom.
How am I supposed to know?
Edit: oh, and I submitted a ticket at 0:46 and the service didn't come back until 1:03 approximately (seventeen minutes). So unless you're saying I predicted the service disruptions...
Negative.
I have a 2 VPSs with Nick each on different nodes. Both display the same amount of downtime... and I wonder why... a main network issue in a rack is going to affect the rack as a whole.
All times below are EDT and in the AM:
12:44:42 - 12:45:37 (timeout)
12:46:03 - 12:46:22 (timeout)
12:48:03 - 12:48:08 (timeout)
12:48:22 - 12:48:28 (timeout)
12:49:37 - 12:49:44 (timeout)
12:50:07 - 12:51:12 (timeout)
12:58:38 - 01:00:47 (timeout)
There you go, down to the seconds. About 4 minutes of real downtime.
Another thread where a user reports downtime on a forum rather than to the host's support, meh.
That's sixteen minutes of network issues. Anyway, I already stated, I had eighteen minutes of network issues, I don't know why you had less downtime but I have three different sources of information (SSH, uptime tracker, services running on the VPS itself) that say the same thing.
Edit: saying that on and off for sixteen minutes is only four minutes of downtime is ignoring the fact that some services require continuous connectivity. I already said that I was considering the entire period of network issues as downtime.
There's no such thing as "real downtime". It depends on what you're running. Anyway your measurement of downtime is completely useless.
So, like I said, there was approximately twenty minutes of downtime.
Edit2: oh, and Wikipedia:
My VPS failed to "provide its primary function" for the entire eighteen minutes, not just the four to six minutes when it was inaccessible. (Not to mention that, I assume because of the switch over, some locations still couldn't connect for another ten minutes+ after).
Lol... what compiled you to bold your statement? I'm curious.
I didn't compile anything, I just used Markdown
Edit: I don't think I was compiled either.
Err, compelled*
Fact of the matter is it was 16 minutes of network instability/packetloss, 4 minutes of real downtime. Don't piss on Nick's company because you think 16 minutes of network inconsistency was downtime.
a) never pissed on RamNode
b) you clearly still have misconception about definition of downtime.
Can someone close this thread now? Network has been stabilized for a while, and the RAID problem is unrelated and only affects 30 people (all of whom have been restored or given new VPSs).