NON-CC GVH

[joepie91]

Bookmarked. Let's see how long this offer will last for.

 

[Francisco]

Dat slabbin'

 

[kaniini]

Dat slabbin'

Yeah I noticed the hypervisor bit was set later.  Going to see if I can find out more.  It's probably KVM I bet.

 

[GVH-Jon]

Someone is hammering the I/O on tx3. The I/O performance should not be that low at all. I'll have everything looked at and resolved shortly, and the node will be more closely monitored.

tx3 is single slabbed with a Xen PV Hypervisor for performance & optimization purposes.

 

[kaniini]

Performance & optimization purposes?  Whaaaaaat?

He isn't lying though -- it is running as a Xen PV hypervisor...


[root@mordor ~]# ./cpuid2
Xen found... checking CPUID leaf 0x40000001
Detected Xen v4.3 [PV]

If anyone wants the code... let me know.  I think I will put this on Github so people can detect if they're being slabbed or not.

 

[hellogoodbye]

Someone is hammering the I/O on tx3. The I/O performance should not be that low at all. I'll have everything looked at and resolved shortly, and the node will be more closely monitored.

tx3 is single slabbed with a Xen PV Hypervisor for performance & optimization purposes.

Oh good! I hope it's fixed soon, I just checked right now and it's still more or less the same:

dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync
16384+0 records in
16384+0 records out
1073741824 bytes (1.1 GB) copied, 84.3402 s, 12.7 MB/s

CPU model : Genuine Intel(R) CPU @ 2.00GHz
Number of cores : 4
CPU frequency : 2000.054 MHz
Total amount of ram : 2024 MB
Total amount of swap : 2024 MB
System uptime : 11:21,
Download speed from CacheFly: 52.0MB/s
Download speed from Coloat, Atlanta GA: 41.7MB/s
Download speed from Softlayer, Dallas, TX: 46.1MB/s
Download speed from Linode, Tokyo, JP: 6.92MB/s
Download speed from i3d.net, Rotterdam, NL: 11.1MB/s
Download speed from Leaseweb, Haarlem, NL: 11.0MB/s
Download speed from Softlayer, Singapore: 5.52MB/s
Download speed from Softlayer, Seattle, WA: 29.9MB/s
Download speed from Softlayer, San Jose, CA: 28.1MB/s
Download speed from Softlayer, Washington, DC: 29.6MB/s
I/O speed : 15.3 MB/s

 

[GVH-Jon]

Performance & optimization purposes?  Whaaaaaat?

The sys admin that set up tx3 couldn't get flashcache set up for some reason so in order to make use of the SSD we have attached to the mobo we decided to single slab the hardware node into the virtualization we found most suitable, Xen PV.

We're still working on getting things worked out though, a lot to do today.

 

[zzrok]

The sys admin that set up tx3 couldn't get flashcache set up for some reason so in order to make use of the SSD we have attached to the mobo we decided to single slab the hardware node into the virtualization we found most suitable, Xen PV.

We're still working on getting things worked out though, a lot to do today.

If you couldn't figure that out, why the hell are you selling the product?  Get your shit in order before you start selling.

 

[kaniini]

The sys admin that set up tx3 couldn't get flashcache set up for some reason so in order to make use of the SSD we have attached to the mobo we decided to single slab the hardware node into the virtualization we found most suitable, Xen PV.

We're still working on getting things worked out though, a lot to do today.

By flashcache do you mean on the RAID card, or is this mdraid + flashcache?

Either way, I can help you do this correctly for a customary fee.

 

[SPINIKR-RO]

By flashcache do you mean on the RAID card, or is this mdraid + flashcache?

Either way, I can help you do this correctly for a customary fee.

Please, he has 18 staff to deal with things like this. Shame trying to belittle Jon with your offering to assist.. psh

Anyways his secretary likely autopened the above post, hes likely beach side by now.

 

[drmike]

By flashcache do you mean on the RAID card, or is this mdraid + flashcache?

Either way, I can help you do this correctly for a customary fee.

I doubt he knows unless the person doing the work conveyed this.  I am certain he could use real technical help as-needed.

Anyways his secretary likely autopened the above post, hes likely beach side by now.

Technically, I think he's in class.  It's a school day

 

[hellogoodbye]

Another update, I/O speeds are looking a little better now!

dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync
16384+0 records in
16384+0 records out
1073741824 bytes (1.1 GB) copied, 10.9726 s, 97.9 MB/s

CPU model :  Genuine Intel(R) CPU  @ 2.00GHz
Number of cores : 4
CPU frequency :  2000.054 MHz
Total amount of ram : 2024 MB
Total amount of swap : 4096 MB
System uptime :   19:55,
Download speed from CacheFly: 50.2MB/s
Download speed from Coloat, Atlanta GA: 43.0MB/s
Download speed from Softlayer, Dallas, TX: 48.0MB/s
Download speed from Linode, Tokyo, JP: 9.04MB/s
Download speed from i3d.net, Rotterdam, NL: 10.8MB/s
Download speed from Leaseweb, Haarlem, NL: 3.24MB/s
Download speed from Softlayer, Singapore: 4.19MB/s
Download speed from Softlayer, Seattle, WA: 19.2MB/s
Download speed from Softlayer, San Jose, CA: 28.0MB/s
Download speed from Softlayer, Washington, DC: 31.5MB/s
I/O speed :  84.4 MB/s

 

[Virtovo]

Another update, I/O speeds are looking a little better now!


dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync
16384+0 records in
16384+0 records out
1073741824 bytes (1.1 GB) copied, 10.9726 s, 97.9 MB/s

CPU model :  Genuine Intel(R) CPU  @ 2.00GHz
Number of cores : 4
CPU frequency :  2000.054 MHz
Total amount of ram : 2024 MB
Total amount of swap : 4096 MB
System uptime :   19:55,
Download speed from CacheFly: 50.2MB/s
Download speed from Coloat, Atlanta GA: 43.0MB/s
Download speed from Softlayer, Dallas, TX: 48.0MB/s
Download speed from Linode, Tokyo, JP: 9.04MB/s
Download speed from i3d.net, Rotterdam, NL: 10.8MB/s
Download speed from Leaseweb, Haarlem, NL: 3.24MB/s
Download speed from Softlayer, Singapore: 4.19MB/s
Download speed from Softlayer, Seattle, WA: 19.2MB/s
Download speed from Softlayer, San Jose, CA: 28.0MB/s
Download speed from Softlayer, Washington, DC: 31.5MB/s
I/O speed :  84.4 MB/s

Looking promising.  Once everyone is finished hammering the node it may rise even further.  

 

[DomainBop]

The sys admin that set up tx3 couldn't get flashcache set up for some reason so in order to make use of the SSD we have attached to the mobo we decided to single slab the hardware node into the virtualization we found most suitable, Xen PV.

We're still working on getting things worked out though, a lot to do today.

A. the sys admin is obviously incompetent.  Did you find him on fiverr?

B. Why did you post an offer before you fixed the server (i.e. why did you intentionally sell your customers a piece of crap)?  Do you think it's fair to your customers that you advertised something as "SSD Cache" when the reality is it performs like a PATA drive?

If you want to build a long-term business that will be sustainable you need to start thinking more of your customers and how to offer them a reliable product, and less about how much $$$ you can make in the shortest period of time.

TL;DR version: look to Prometeus/RamNode, etc for inspiration on how to run a hosting business rather than trying to emulate Fabozo and his crew of circus performers.

 

[Hxxx]

A. the sys admin is obviously incompetent.  Did you find him on fiverr?

B. Why did you post an offer before you fixed the server (i.e. why did you intentionally sell your customers a piece of crap)?  Do you think it's fair to your customers that you advertised something as "SSD Cache" when the reality is it performs like a PATA drive?

If you want to build a long-term business that will be sustainable you need to start thinking more of your customers and how to offer them a reliable product, and less about how much $$$ you can make in the shortest period of time.

TL;DR version: look to Prometeus/RamNode, etc for inspiration on how to run a hosting business rather than trying to emulate Fabozo and his crew of circus performers.

This above comment is so dumb. You always have to throw the same shit?

 

[drmike]

B. Why did you post an offer before you fixed the server (i.e. why did you intentionally sell your customers a piece of crap)?  Do you think it's fair to your customers that you advertised something as "SSD Cache" when the reality is it performs like a PATA drive?

This is a good summary and lesson for GVH-Jon.  Buy server, config and install, load test... test some more... pound the network a bit... reboot, retest...  thrash disk for a bit... repeat..

3 days later, maybe you can have confidence in selling on that server.  

It's a shame that his first non-CC offer underimpressed up until to this point.  SSD cache + RAID controller = not free.  But the numbers have gone up, which good.

This above comment is so dumb. You always have to throw the same shit?

I resemble that comment.

 

[Virtovo]

A. the sys admin is obviously incompetent.  Did you find him on fiverr?

B. Why did you post an offer before you fixed the server (i.e. why did you intentionally sell your customers a piece of crap)?  Do you think it's fair to your customers that you advertised something as "SSD Cache" when the reality is it performs like a PATA drive?

If you want to build a long-term business that will be sustainable you need to start thinking more of your customers and how to offer them a reliable product, and less about how much $$$ you can make in the shortest period of time.

TL;DR version: look to Prometeus/RamNode, etc for inspiration on how to run a hosting business rather than trying to emulate Fabozo and his crew of circus performers.

Exactly.

 

[D. Strout]

Anyone have one of this offer with working IPv6? I didn't get any allocated at first. I ticketed and got them (along with one of the 25 free year + IPs slots), but they don't work. ping6, traceroute6, and wget -6 all fail. I sent in another ticket, but I'm curious if anyone else is having this issue.

 

[Kadar]

Greetings Valued Clients,

 

This email is being sent out to you today to notify you that we will be upgrading our srv3 clients and moving all clients on srv3 to a brand new pure SSD server in Chicago, Illinois, USA (Central). We understand that there has been recent issues with the performance on srv3 even after the upgrades that have been made on it and we've decided to move all srv3 clients (And later on, srv5 clients) to pure SSD servers in Chicago, IL as our Buffalo, NY upstream cloud hosting provider has decided to be uncooperative with us.

 

srv3 clients will be moved to a server with the following specifications:

 

Intel Xeon E5-2620 Processor (6 cores physical, 12 logical cores)

8x 128GB Intel SSD Drives in LSI Hardware RAID-10 w/ Battery Backup Unit

32GB DDR3 ECC RAM

 

This server unfortunately will not be in the cloud (We're going to be discontinuing our offerings of cloud hosting in all locations with the exception of Los Angeles, and deploy Pure SSD servers across all the locations we offer), however though we are confident and will guarantee you that the performance of this new server will be far more optimized and ideal than the current server.

 

The migration is scheduled to take place on Saturday, Februrary 1, 2014 and we will be sending out an email to notify you when the migration begins. As there are a large amount of accounts on srv3, the process may take a while (possibly a few days). Further information will be included in the upcoming email to be sent out before the migration begins.

 

If anyone has any questions regarding this, please let us know by ticketing in to our customer service department and we'll always be more than happy to help.

 

Kind Regards,

 

The GreenValueHost Team

 

[SPINIKR-RO]

heh, well at least he is not selling the customers off again. 

Not really sure why the upstream provider has to do with not being able to fix a issue with the server. I assume all of the "cloud" stuff is a big VM provided by the upstream and GVH was looking to them to fix whatever the servers issue was.

We're going to be discontinuing our offerings of cloud hosting in all locations with the exception of Los Angeles

Just to be clear, the entire argument as to why GVH sold their customers was to move everything to the 'cloud' and now no more cloud!