Just finished a good talk with Zak on Skype. Very satisfied with the outcome. He came up with a good plan and upgrades to keep the issue from happening again. More than made up for the issue and then some.


Thanks to Brad for his feedback on the issue too. Wish there were more knowledgeable hosts posting on GFY.

:pimp Good to hear about a positive outcome :thumbsup
These things happen, and however it's very hard to accept the downtime you Must deal with it. Downtimes always end and you can focus on working and usually there are No big losses. A month later you won't think about this x hour downtime as something that barely killed your business, but something that was fucking annoying but is OVER. :2 cents:

Peace

Yes I have. And no, I never went against his advice. If I knew how to manage hosting, I wouldn't need managed hosting, would I?

The mind is strong, the flesh is weak... :Oh crap

I do my best to avoid getting into technical debates with other sysadmins, really. It's a complete waste of time but for some reason I'm living dangerously this morning...

If I had a media client w/ 6TB of storage I'd do whatever it takes to get them off ext3 if for no other reason than the probability they are at or near the folder count limit for the file system.

And, mandatory FSCK is NOT required and can be disabled, especially if you have a nice, replicated SAN or NAS storage system where you can rebuild file system errors while the system is online. Do you seriously believe Enterprise clients on Linux are willing to sit around and wait hours on end for a fsck to complete after a crash>? I spent most of the past decade designing and installing storage solutions for the largest corporations in the World and I've never seen it happen (and no, they were not all proprietary SAN or NAS).

Not to mention the fact that most adult hosting servers are not writing data to the file system that requires such drastic measures, the worst you might see would be corruption in the database which you will quickly find out about and that can be fixed in a few minutes if it occurs. Most adult sites are only writing critical data to the database, everything else is temporary minus the rare occasion when a site edit or encoding session is underway at the time of the crash. Most clients would probably re-do that last task vs. be offline for several hours.

Of course, ditching R1 as your backup solution goes without saying in this scenario. Then again, any client paying me 1k plus per month would be on a load balanced, highly available architecture anyway. It looks like he'll be paying 2k in the near future though and I can't fault any provider for getting some, if you can charge that much for what you give him I envy your sales abilities / incredible luck in getting clients like that.

Dreamhost!!

Yes, that's why you probably don't know what i, and others are talking about. It's not my place to question how you do your business, but zak is not all knowledgeable, believe it or not, and there's things he's really clueless about. But, he acts like he knows everything about anything. So once there's a difference of opinions, he puts up an attitude and is impossible to work with.

Great post. You should write more often.

Mojohost forever. (knocks on wood)

wow :disgust :helpme

Yep there is, and I can understand how this is. Everyone get off his back. He's doing the best he can.

Ok, so....

1. Zak's an asshole.
2. He warned you against doing something or another.
3. Whatever he warned you against happened.
4. He is now being an asshole saying I told you so.

Is that the thread in a nutshell? Well,... except for the sabre rattling, and legal mumbo jumbo.

:helpme

Pretty far off.. on a lot of parts.. but in a nutshell..

Shit happens
shit was resolved
shit was improved and made better free of charge
Zak may or may not still be an asshole. Up to personal interpretation.

ouch file system checks are grueling but no hosting company is gonna be able to make those go any faster especially on that much data!

cliffnotes

yellow fiber is a good host and the ghost had an irrational tantrum

:thumbsup

Glad to see you guys got shit resolved :thumbsup

True dat fo shizzie.

:pimp

I love your bandwidth, your network, and how quickly you guys reply to my tickets. I do manage my own servers and fix my own problems.

Something you might want to look in to for this clown of a customer is ZFS. I use it on 12TB+ storage arrays. Love it. Would never go back, and the incremental snapshots are a huge plus!

I always thought their support was good since they got those russian guys at Far Post on board doing it?

Webair FTW....


19:37:32 up 414 days, 1:26, 1 user, load average: 0.00, 0.00, 0.00

mojohost

Sounds like you're due for a fuck-up. :1orglaugh

How does the power not going out for a year equate to a good host? Could have had hours of network related downtime (which is more common) over that period.

Shouldn't they be updating your kernel and rebooting it every now and then?

Zack should spend more time assisting clients rather than dissing the competition