All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
HardCloud "SSD-Caching related failure"
Just got an e-mail from HardCloud
Hello,
Your service with the IP XXX.XXX.XXX.XXX and hostname XXXXXX may be at risk of data loss! If your service is located on "Slave4" please open a ticket in our blling portal to facilitate a data-transfer.
Slave4 has recently recovered from an SSD-Caching related failure, one o the cache drives failed and we've lost the configuration files for osme virtual machines on this server. The Virtual Machines are still functioning now on the backup array, but they will not function here for long. After 24 hours, we will terminate all Virtual Private Servers left on Slave4 and re-image it, then re-deploy all Virtual Private Servers located on it.
Please backup your data or open a ticket to have us backup your data immediately, otherwise you may lose it.
Yours sincerely,
HardCloud Staff
Comments
Yeah... got that.
Seems to me they could just back up everyones data and save all the ticket drama. I'm pretty sure no one wants their data deleted so it's a bit of a non-choice.
Or just live migrate everyone to a new node, before this one falls apart completely? People wouldn't even notice it.
I am so confused, how does a SSD-Caching that fail cause 100% data loss?
Maybe they used Writeback on the cache to boost performance at the cost of data safety.
But if they recovered, can't they simply disable caching and pretend nothing happened?
"We screwed up please request that we fix it."
I thought data loss was impossible on hard cloud because the backups are always on the mirrors?
Those typos bug me.
Same.
Also, is this the same provider who believes that 2 CPU cores = failover?
well it sounds like they did not really recover, if you have Write back on the cache it is written to the SSD first then the storage after which obviously gives better performance but if you have issues with either the SSD's or the storage shit gets real, your options are to turn off the cache but possibly with missing data depending on the severity and time it took to notice.
Still,
The dataloss won't be major. After all, it will endup writing back to the original HD.
Maybe SSD caching is their "marketing phrase" for RAID 0? :P
amusing...
How do you know if your on slave4 or not ?
SolusVM
****edited****
nevermind
Hello,
Yes; the CacheCade (2x SSDs in RAID1) Virtual Disk failed on our node, as it was improperly shut down by the data-center and was using WriteBack (albeit, with the BBU, but in the case of an improper shutdown it is a non-issue.) We have re-deployed all VPSes that were affected by this already.
Please note that most of the comments portrayed by the "HardCloud" as quoted in this thread are those of the previous management and owner of HardCloud, not the current management and owner of HardCloud. This is a non-issue that has been solved elsewhere on LowEndTalk and verified by moderators.
Q) A BBU will do what exactly with writeback enabled during an in-proper shutdown?
A) Loose data
Anyway, best of luck.
Sure...
His comments are just as terrible so I don't know why we're splitting hairs over which incompetent employee/tech posts them?
Can you confirm that? My server is still down and solus has the server status as "unknown".
Hello Ricardo,
Feel free to open a Support Ticket in the panel and we will look into this.
Hello Critics,
I have already had a moderator confirm that the previous owner went a bit mad before handing over the marketing accounts. That's how the MAX_VPS post and related articles have been deleted, and the HardCloud Failures thread now says "Previous HardCloud Failures".
I'd like it if my posts weren't mis-quoted either, thanks.
It's kind of hard for people to move past it. Let's look at the hard(cloud) facts: someone in your company, claiming to be in a leadership position, was using the company's public account to bad-mouth another provider's post. Then found out to be misinformed, and blatantly lying to potential customers. Pretty bad wrap if you ask me.
HA!
Also, I'm not sure if you're aware but there's a WHMCS exploit just released, you should shut down your client system to prevent a compromise in customer data.
They don't need to they have a BBU which will be just as effective if it gets in-properly accessed.
Hello,
It seems you like to ignore what I type, specifically:
That's rather annoying and counter-productive to both of us. Why can't we all just get along?
Thanks for the heads up, looks like mod_security CRS won't protect us this time.
.... sure, I just need to repair my logic engine, you broke it.
I would but WHMCS is down, understandable given the exploit.
No hurry actually. I can wait a day or two and will create the ticket then if it persists.
Quickly sorted. Thank you.