oles@ovh.net
11-02-2008, 07:56 PM
Hello,
Here is the current status of deliveries (2008-11-02 20h00).
Real Private Server
- RPS 1 OK -> 1 hour
- RPS 2 7 days delay (The oldest order paid and not deployed is dated 2008-10-27 21:25:10)
- RPS 3 4 days delay (The oldest order paid and not deployed is dated 2008-10-30 08:51:05)
Servers to experiment and to play with:
- Kimsufi L OK -> 1 hour
- Kimsufi XL OK -> 1 hour
- Kimsufi 2XL OK -> 1 hour
- Kimsufi 3XL OK -> 1 hour
- Kimsufi 4XL OK -> 1 hour
Pro, business and critical servers:
- miniSP OK -> 1 hour
- SP bestof OK -> 1 hour
- SP Max fr,es OK -> 1 hour
- SP Max de OK -> 1 hour
- SP Max uk,pl OK -> 1 hour
- EG Bestof OK -> 1 hour
- EG Max OK -> 1 hour
- EG Large OK -> 1 hour
- EG SSD OK -> 1 hour
- MG Bestof OK -> 1 hour
- MG Max OK -> 1 hour
- MG Large OK -> 1 hour
- MG SSD OK -> 1 hour
- MG X Large OK -> 1 hour
- HG 2009 (being prepared)
We still have difficulty in controlling the growth of RPS. We are increasing production every week, but a lot of you are ordering RPS and the stock empties very fast. We'll change the strategy and this week we will make available what we have planned for 2 weeks. We'll see if it's good enough or not.
The rest is delivered without any problems. We had a minor backlog last week on SSD servers but the delay was caught up before the end of the week. Indeed, sales of SSD servers were not successful the 1st week. Many clients have had bad experiments with SSD drives. But the noise on the excellent performance of Intel SSD drives quickly spread around the Internet. For example Linus Torvalds (the father of Linux) has also tested the Intel SSD disc (the Intel X25-M) and was knocked out when he saw the performance. [http://torvalds-family.blogspot.com/2008/10/so-i-got-one-of-new-intel-ssds.html As a result, according to our feedback from Intel, the whole "integrator" market has woke up this week and hopes to soon integrate this new Intel baby. At our level, X25-M intel disks are available in the EG and MG range and despite the fact that Intel now sees SSD drives sales growing, we should ensure the availability of these servers in 1 hour.
The new range of HG 2009 arrives in a few days, probably this week. We are in the home straight before the creation of the site and the establishment of the order. Everything else is already available and servers have already been manufactured. The servers are already operating in the racks and responding to ping ... We received all the parts, including Intel SSD drives Intel X25-E ... the PRO brother of Intel X25-M, which is even more faster, not that much, just 3-4 times faster but it takes ... Intel X25-E will be available in the new Sun servers and ... at Ovh this week. Meanwhile, it is frustrating to have everything and not be able to announce to you the HG range, but also to innovate the hardware side (SSD is not the only surprise in the new HG 2009), we whish to present to you a new vision for the presentation of the lease of dedicated servers. In contrast to the 2008 range, in the full 2009 range, it means SP, EG, the MG and HG, we wish to offer you maximum customization of everything, the best possible price and availability in 1 hour. That is our challenge in 2009: maximum customization while ensuring 1 hour availability. We will start modestly with the HG 2009 range ... this week.
For 10 days, we set up the Standard and Premium SLA we announced in August / September. Indeed, in the "revolution" range we have announced the establishment of 2 qualities of network: for Kimsufi and for PRO servers. We left 2 months to our customers to take note of these changes and now we check the real use of bandwidth of your servers and we apply the Standard and Premium SLA. In total, up to about 80/90 today have been affected by the change of the quality of the bandwidth following breach of the contract. 80/90 servers out of 42,000 servers it's 0.2% of our customers but we wish to remain transparent as much as possible on this implementation. That's why we provide the technical explanations on "How the SLA and the SLA Standard Premium works ": http://translate.google.com/translate?u=http%3A%2F%2Ftravaux.ovh.com%2F%3Fdo%3 Ddetails%26id%3D2495&hl=en&ie=UTF-8&sl=fr&tl=en with the technical parameters, and transparency on the establishment of QoS rules on our network.
This week, we set up the Business SLA that you can put in place on your servers and up to 500Mbps of guaranteed bandwidth per server! Ovh is currently negotiating several large contracts with some large radios on the Internet in Europe.
We have fixed the problem on the 10G on Linx 224 in London. The problem was on Linx'side. After changing the port 10G and optics, the problem has been resolved. I wanted to thank personally UK customers who helped us in the identification of the problem and have assisted us in the tests. It is not easy to find the source of a problem when it's not on our network ...
This week we will switch Decix in Frankfurt on the router directly. Indeed, load balancing of 10G links in a port channel is not homogeneous on each 10G port on Force10 equipment and we'll switch to Cisco. We have the same problem on AMSIX. You can see on our weathermap on Decix and AMSIX, the 2 10G links do not have the same load http://weathermap.ovh.net/backbone.
Regards,
Octave
Here is the current status of deliveries (2008-11-02 20h00).
Real Private Server
- RPS 1 OK -> 1 hour
- RPS 2 7 days delay (The oldest order paid and not deployed is dated 2008-10-27 21:25:10)
- RPS 3 4 days delay (The oldest order paid and not deployed is dated 2008-10-30 08:51:05)
Servers to experiment and to play with:
- Kimsufi L OK -> 1 hour
- Kimsufi XL OK -> 1 hour
- Kimsufi 2XL OK -> 1 hour
- Kimsufi 3XL OK -> 1 hour
- Kimsufi 4XL OK -> 1 hour
Pro, business and critical servers:
- miniSP OK -> 1 hour
- SP bestof OK -> 1 hour
- SP Max fr,es OK -> 1 hour
- SP Max de OK -> 1 hour
- SP Max uk,pl OK -> 1 hour
- EG Bestof OK -> 1 hour
- EG Max OK -> 1 hour
- EG Large OK -> 1 hour
- EG SSD OK -> 1 hour
- MG Bestof OK -> 1 hour
- MG Max OK -> 1 hour
- MG Large OK -> 1 hour
- MG SSD OK -> 1 hour
- MG X Large OK -> 1 hour
- HG 2009 (being prepared)
We still have difficulty in controlling the growth of RPS. We are increasing production every week, but a lot of you are ordering RPS and the stock empties very fast. We'll change the strategy and this week we will make available what we have planned for 2 weeks. We'll see if it's good enough or not.
The rest is delivered without any problems. We had a minor backlog last week on SSD servers but the delay was caught up before the end of the week. Indeed, sales of SSD servers were not successful the 1st week. Many clients have had bad experiments with SSD drives. But the noise on the excellent performance of Intel SSD drives quickly spread around the Internet. For example Linus Torvalds (the father of Linux) has also tested the Intel SSD disc (the Intel X25-M) and was knocked out when he saw the performance. [http://torvalds-family.blogspot.com/2008/10/so-i-got-one-of-new-intel-ssds.html As a result, according to our feedback from Intel, the whole "integrator" market has woke up this week and hopes to soon integrate this new Intel baby. At our level, X25-M intel disks are available in the EG and MG range and despite the fact that Intel now sees SSD drives sales growing, we should ensure the availability of these servers in 1 hour.
The new range of HG 2009 arrives in a few days, probably this week. We are in the home straight before the creation of the site and the establishment of the order. Everything else is already available and servers have already been manufactured. The servers are already operating in the racks and responding to ping ... We received all the parts, including Intel SSD drives Intel X25-E ... the PRO brother of Intel X25-M, which is even more faster, not that much, just 3-4 times faster but it takes ... Intel X25-E will be available in the new Sun servers and ... at Ovh this week. Meanwhile, it is frustrating to have everything and not be able to announce to you the HG range, but also to innovate the hardware side (SSD is not the only surprise in the new HG 2009), we whish to present to you a new vision for the presentation of the lease of dedicated servers. In contrast to the 2008 range, in the full 2009 range, it means SP, EG, the MG and HG, we wish to offer you maximum customization of everything, the best possible price and availability in 1 hour. That is our challenge in 2009: maximum customization while ensuring 1 hour availability. We will start modestly with the HG 2009 range ... this week.
For 10 days, we set up the Standard and Premium SLA we announced in August / September. Indeed, in the "revolution" range we have announced the establishment of 2 qualities of network: for Kimsufi and for PRO servers. We left 2 months to our customers to take note of these changes and now we check the real use of bandwidth of your servers and we apply the Standard and Premium SLA. In total, up to about 80/90 today have been affected by the change of the quality of the bandwidth following breach of the contract. 80/90 servers out of 42,000 servers it's 0.2% of our customers but we wish to remain transparent as much as possible on this implementation. That's why we provide the technical explanations on "How the SLA and the SLA Standard Premium works ": http://translate.google.com/translate?u=http%3A%2F%2Ftravaux.ovh.com%2F%3Fdo%3 Ddetails%26id%3D2495&hl=en&ie=UTF-8&sl=fr&tl=en with the technical parameters, and transparency on the establishment of QoS rules on our network.
This week, we set up the Business SLA that you can put in place on your servers and up to 500Mbps of guaranteed bandwidth per server! Ovh is currently negotiating several large contracts with some large radios on the Internet in Europe.
We have fixed the problem on the 10G on Linx 224 in London. The problem was on Linx'side. After changing the port 10G and optics, the problem has been resolved. I wanted to thank personally UK customers who helped us in the identification of the problem and have assisted us in the tests. It is not easy to find the source of a problem when it's not on our network ...
This week we will switch Decix in Frankfurt on the router directly. Indeed, load balancing of 10G links in a port channel is not homogeneous on each 10G port on Force10 equipment and we'll switch to Cisco. We have the same problem on AMSIX. You can see on our weathermap on Decix and AMSIX, the 2 10G links do not have the same load http://weathermap.ovh.net/backbone.
Regards,
Octave