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
@Spencer, you are a hoot.
A+B power redundancy is kind of mandatory at some size. BuyVM regardless of staff or even financials is a big enough public facing company to mandate A+B power (anyone notice over providers in Vegas getting poked about the outage?). The ATS PDUs are relatively affordable. Probably can find a way to make the ATS units themselves redundant as well, although won't be as affordable.
As for the switching, @Francisco need hot spare in place if not live-live system. That's a big poof piece at any time that goes broken.
I am still impressed everything runs as well as it does. Regardless of the outages.
I'll be saying my special quiet time words hoping Fiberhub gets it together and this makes the last snafu for a while.
A free month? Hot dog!
aka prayer? Or do you just speak in quiet sometimes?
popular enough. household name?
Literally, you can buy a hot dog with what BuyVM charges monthly!
We have a free month for the offended.
@Nick_A, just reflective time. I appreciate my quiet time when lucky enough to get some of that rareness. Just sending positive energy in the needed direction.
@jcaleb, big enough, popular enough. Yes. Certainly the big daddy of this market right here.
Shoot, more like 20 hot dogs.
Those are some cheap hotdogs @Nick_A. You better eat better than that. This industry needs you.
You must be buying those horse meat 'dogs from the UK!
Heard the original problem was a french company that tried to blame a romanian one (why not, everything bad in EU is Romania's fault :P) then their dutch partners seem to have been involved and UK maybe got it from the dutch
At least the horses dont have http://en.wikipedia.org/wiki/Creutzfeldt–Jakob_disease
Yeah, there seems to be a convoluted supply chain from what I have read.
Don't worry, tons of us in the US probably eating horse meat too though. Aldi grocery was implicated in the horse mess and claimed it was their suppliers fault.
Now mind you, years back I went into Aldi and bought some canned fruit. Ends up that lovely product came from China. Not really where I want my food coming from.
LET goes lowend dining
Can you please kick up the shared SQL? This is sort of getting ridiculous.
It goes down every time power goes out, and stays down for 5-6 hours.
Edit: Seems to be back up now, hm.
SQL was up before Anthony had nodes spinning containers.
Since the we didn't have innodb issues this time it was very quick.
I'll see how much power we have spare in a few weeks and I may setup and HA.
Francisco
@pubcrawler, @Microlinux Wait.. what's wrong with eating horse meat?
So glad I picked the NY location.
Got SLA credit. Thanks boss
@Francisco: I'm not going to claim SLA, you'd better invest that dollar on the PDU
Down again before 8 hours!
My database crashed ... again!
130215 14:34:40 InnoDB: Error: page 461 log sequence number 10364237387
InnoDB: is in the future! Current system log sequence number 10364234948.
InnoDB: Your database may be corrupt or you may have copied the InnoDB
InnoDB: tablespace but not the InnoDB log files. See
InnoDB: file read of page 531.
InnoDB: You may have to recover from a backup.
130215 14:34:40 InnoDB: Page dump in ascii and hex (16384 bytes):
I'd have put up with EGi's network any day rather than this.
FH has been extremely abysmal so far.
@Zeno offloaded mysql is meh... buy a few LEB's in the same geo and build a small cluster... that way when outages happen you 1) aren't impacted and 2) can gracefully recover..
Their server doesn't have UPS?
Why are so frequent shutdown?
I dont know, buyvmstatus.com shows green all over.
Offload MySQL is double edged. It is great for enhancing performance but another failure point.
I have a Mysql Slave on another VPS,
But crashes every time I need to export / import 10G data ...
It's so sad .........................
Can you see the NetUp time just 9h 11m?
storage-lv-04 9h 11m
@Zeno it's too bad, sorry. Setup 3 node cluster and let it fail to another master, then resync the failed node..
E.g. if you have 20A with 16A usable on A, you still get the 16A usable on A+B, but it might be 8A on A and 8A on B but max is 16A on A or 16A on B at one time.
In this case it'll be the full 16A on either the A or B feed at any one time. The load won't be balanced over both feeds because they aren't using dual power supplies.
I would also get more than 1 ATS per rack that way if 1 fails it won't take out your whole rack. It will also be switching a lighter current load which may help prolong life.
There is UPS' onsite for the datacenter. We don't run our own since it's costly and a fairly big issue. The problem they had was between their UPS and their distribution so no amount of UPS/generators they had would have fixed it, at least in the current setup would have helped.
Only until coresite had their own power outage. Coresite SJ has had issues in the past involving power failures. Not many but as FH has shown, "it's only a matter of time".
At least now we're able to get A+B as it hasn't been an option to date.
For some people SJ just fine and there was no speed issues due to the time of day they used the network, etc. I had 5+ pages of speed complaints, though, that would rather we fix the problem.
I'm not trying to justify this or anything. trust me, I'm beyond pissed and so is FH themselves since their spotless record is borked now. The issues will be addressed and customers will gain a rarely executer layer of redundancy in an LE provider.
Francisco
thanks @unused
But even so, I also need to spend time to restore data on crashed InnoDB mysql, right?
I hope BuyVM equipment UPS on their servers, do not frequent shutdown