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
MarkTurner Will the new offer have low-end server around $30/m?
me too
If you are still down, then open a ticket so they can investigate.
network outage already exceed 5 hours, and still not recovered, so I had to sent a ticket just now:(
To bad there isn't a SLA..
@hotsnow - the network part of this closed two hours ago. If you are still having issues you must open a ticket so they can investigate. You'll probably need a hard reboot on your server
it's finally online now...
I wonder what actually happened, the route outage and server's hardware issue have occurred on the same time? the route outage shouldn't cause the server crash
same here.. clients almost killed me..
Yeah.. what happened hardware-wise?
I got back online a few minutes ago
Apparently my server needed a cold boot.
That misbehaving router must have been spewing complete garbage, I've never had network problems completely crash a server before.
file a ticket requesting an RCA?
I know they are awsome
I know i was in the middle of a database migration before i switched over to the hardware here it would have been a bad first day if it finished.
We're waiting anxiously for the post-mortem, when I see it I'll summarise on here.
From the network perspective, we upgraded a Juniper router on Sunday with a newer code release. We have got this same version running in Europe and it has been robust. About 11:30PM yesterday we saw a pile of OSPF sessions and then this router establishing and then dropping the session with the other OSPF speakers. With each flap, the routing table got updated and by the time this was happening almost as fast as the session came up, we started seeing packet loss.
As the problem cascaded, it resulted in our routes being internally dampened and also by some of our peers and transit providers. At which point some routes disappeared from the routing table.
The router was finally taken out of service whilst we cleaned up the routing and also re-added the filtering that we had removed whilst we merge some of our east coast network. Router was brought back into service and connectivity was restored.
At that point, its not clear what happened with some customers servers. This was handled locally, so we are waiting for their explanation.
I'm curious to know the full story as well, hope they manage to figure out what happened.
Thanks for getting back to us on this it does mean a lot.
Same here i have never really seen dedicated servers turning them selves off because of some network loss. The other thing is the ikvm/ilo went down with it.
@wojons - The ILO would have been disrupted because it was amongst the netblocks that were off air during the outage.
I don't think the servers turned themselves off, I think the local blade management system saw a loss of service so scheduled a reboot (This is supposition (gut feeling) right now - I have not seen any evidence to support this idea so don't hold it up as an official answer.But its the only way I can explain that out of 2500 blades about 70 'spontaneously' were power cycled). There were about 10 reboot requests made by customers, those are accounted for separately because there was an audit trail.
I will know more tomorrow. According to our network monitoring it was concluded in about 1.5 hours. According to Wormly (who we use for external monitoring): Estimated Downtime: 1 hour 46 minutes 49 seconds
Anyway - how are things for everyone today?
Intresting very intresting. I am having ping issue with a node on your network. Here is something i posted on it.
http://lowendtalk.com/discussion/20486/werid-ping-times#latest
@wojons - PM me the details inc your server and the destination you are trying to reach
Sweet, got my 5150 today.
That's some nice results.
all the things seems normal
@MarkTurner Thanks for the updates as said previously it means a lot and shows you guys care enough to let the customers know. Everything seems normal today.
@Jarland - Good
@hotsnow - Thats what I prefer
@kajak - we try to be as transparent as possible
Patiently waiting on my x5150 to get delivered :-), at least I didn't have to worry about not being able to reach the server I don't have yet haha.
Thanks for keeping everyone updated Mark!
Just finished the serverbear. Good comparison for this thread. Straight out of the box, Ubuntu usually outperforms CentOS. Plenty of ways to increase performance, but that's just my experience right out of the box on most systems.
http://serverbear.com/benchmark/2014/01/23/PSmqI3etlhdKkdx6
everything is normal. clients happy, i am happy, delimiter will be happy too when it comes to next billing, in short, everyone's happy.
While I do applaud your efforts here, I have to point out that's a rather ironic comment considering Yomura is one of the most intentionally shadowy and elusive companies (or "companies" - are you even a legally registered entity?) in existence.
@dylan - Delimiter and Yomura are both legally registered companies both in Europe and US. I think Delimiter put their legal entity information on their website.
As for Yomura, you are spot on about its 'stealth mode' mentality. Yomura has always kept itself to itself, avoiding the press-releases, the limelight on a corporate level. Many reasons have been put forward for that, but they all boil down to a mentality type. They want their product/brands to speak for themselves.
Yomura at the end of the day is just the 'bank' for these companies. They fund them and give them facilities/services to sell. The brands are customer facing and handle the sale/support. Yomura just wants to sit in the background and receive its profits.
If you read my wording 'transparent as possible' - I mean those words. Within the boundaries created by the company's extensive constraints, we (in this case I) try to be as transparent as I can within that framework.
@Jarland - looks like you have one of our 'Jamaica' servers Looks like Serverbear's IP to Geo database is wrong again
correct me if I'm wrong but on windows it shows up as 2 processors so that means 2 cores? so how come I don't get 4 cores