OVH Community, your new community space.

Deliveries 2009-03-14


oles@ovh.net
17-03-2009, 11:36
derchris a écrit:
>
> Regarding the "Bugs" for the KVM switches.
> What "Bugs" are we talking about?


- keyword encoding is lost after a hard reboot of the server
- sharemedia doesn't work before booting

derchris
14-03-2009, 15:46
Regarding the "Bugs" for the KVM switches.
What "Bugs" are we talking about?

I mean, we are using about 100+ KVM switches at work, and never had a problem with them.

derchris
14-03-2009, 15:43
Hello,
Here the delivery condition to 2009-03-14 12:00.

When there is a delay, it is necessary to read:
“the oldest order paid and not carried out has XX hours of delay”

Real Private Server
- RPS 1 8J (2009-03-06 12:28: 10)
- RPS 2 OK - > 1 hour
- RPS 3 6J (2009-03-08 22:27: 06)
- RPS 4 8J (2009-03-06 14:53: 05)
- RPS 5 5J (2009-03-09 15:15: 08)

Waiters to discover and of plays:
- Kimsufi L OK - > 1 hour
- Kimsufi XL 4J (2009-03-10 22:35: 11)
- Kimsufi 2XL OK - > 1 hour
- Kimsufi 3XL OK - > 1 hour
- Kimsufi 4XL OK - > 1 hour
- Kimsufi 8XL OK - > 1 hour

Waiters of company, professionals and critical:
- miniSP 2J (2009-03-12 20:00: 29)
- SP bestof OK - > 1 hour
- SP max Fr, be OK - > 1 hour
- SP max of OK - > 1 hour
- SP max the U.K., pl OK - > 1 hour
- SP Storage OK - > 1 hour

- EG Bestof 2J (2009-03-12 20:03: 15)
- EG AMD OK - > 1 hour
- EG max 2J (2009-03-12 18:20: 16)
- EG Large OK - > 1 hour
- EG SSD OK - > 1 hour


- MG Bestof OK - > 1 hour
- MG AMD 10J (2009-03-05 23:15: 08)
- MG max OK - > 1 hour
- Broad MG OK - > 1 hour
- MG SSD OK - > 1 hour
- MG X Broad OK - > 1 hour

- HG 2009 BestOf OK - > 1 hour
- HG 2009 BiTurbo OK - > 1 hour
- HG 2009 SSD 48H (2009-03-13 14:11: 05)
- HG 2009 Broad OK - > 1 hour
- HG 2009 Personalized OK - > 24 hours

Since the changes on the range of RPS, we have a little evil
to follow the rate of your orders. Since the product is
stable and has many advantages, we record up to 200
orders per day (not every day, but nevertheless). Expenses
of installation on Kimsufi are there more something too. Us
thus let us work on the improvement of delivery period. Approximately, it
acts to see on 10/14j the number of orders and to fix the production
with the number of RPS that you order us. It is thus necessary for each
time to give us a little time so that one can see what it is necessary
to produce and the temporary production of manner is changed or
final. We have also some bugs still to fix in
procedures of change of RPS. Indeed, very few customers make
the evolutions of RPS (that proves well that Cloud Computing it is one
virtual concept) and one finds some bugs in 1 changes on
10 qu ' one will have fixed a whole day. In any case, you seem to adore
Atom < 10Euro and 4 cores for < 40Euro!

On the level of Kimsufi, since the installation of the new XL, it is
the madness on Kimsufi XL. One makes our possible and I think that
running the week one will have to stabilize the production well.
It to remain us to fix the delivery of Kimsufi XL with Windows Pro.
That will be done this week.

You order us much EG BestOf and Max. For the moment
one is not able yet to produce these waiters in the quantities
sufficient by day. With the starting of the 4th production line
waiters, this problem should disappear. The installation of
this 4th line was decided 4 weeks ago and we seek
always new collaborators to reinforce our teams of
manufacture of waiters. One nicely will say that the Pole of Employment
(in the past ANPE) many problems with our advertisements cause us
of looking for a job. The formation is longe and the dependent investment
with this formation very heavy. We thus seek the collaborators
“stable/for a longe durée/CDI” and not “disposable”. Blow our
advertisements of employment are filtered by people of Pole Of employment which
the candidates prefiltrent that they send to us and with the finale one starts
to think that it is pains it employment in Roubaix-City… we thus have
started to communicate on the recruitings which we realize to have
candidates… It is sad but one seeks people and one has an evil
insane to make publicity necessary of it to receive the candidates
by the networks which are there for that. More: http://www.ovh.com/jobs.

We have engineering problems on MG AMD with an incompatibility
between the graphics card and the distributions. One should fix it
problem running of the week and to deliver all the waiters late.

We will increase HG stocks in order to have at least 10
waiters of each reference in advance. This makes it possible to start them
large infrastructure in a few minutes.

As already announced, after the IP LoadBalancing, we work on
VLAN in version “not undertaken and “undertaken”. In the same way, the VPN and
the firewall. We still have problems on the KVM which us
do not give satisfaction. One wonders how our competitors
can propose the KVM with as many bugs and what the customer
must say when the KVM is cut when one reboote the waiter or when
the configuration of the KVM is erased. One does not include/understand. Thus for
the moment, one goes up the bugs with the manufacturers so that they fix them
bugs. If it were opensource that would be already corrected… Bon it
KVM one continues the tests and the fixing of bugs.

The equipment for the network Frankfurt/Brussels/Amsterdam/Roubaix
left the USA and arrive in the week (weekend)
in Amsterdam. We will deploy any weekend, the weekend
and beginning of the week. One should have any operation in 10 days
(yes!). The equipment for the network Roubaix/London arrives
10 days after and one deploys it in pressed too. In pressed
one connects new the 10G transit with Teleglobe and Global
Crossing, one rebalances the network and one push 200/220Gbps at the beginning of April.
(yes!). Then one adds the new routers with… for… (yes!)

The configuration of peering with Orange and TPSA are programmed for
Tuesday morning. One should start in pressed, but the peering with
TPSA with Frankfurt will be completely used once the network until
Frankfurt is finished.

A new router VSS arrived and will be set up in the days
who come. Indeed, the current one will be fills under some 2-3 weeks.
There remains to us the foutu bug in the IOS of our VSS to be fixed then one is
good on the routing. What will enable us to look for the network with
Pt, SE, NO, DK… There is taf

In a friendly way
Octave

oles@ovh.net
14-03-2009, 12:34
Hello,

Here is the current status of deliveries (2009-03-14 12h00).

Real Private Server
- RPS 1 8 days (The oldest order paid and not deployed is dated 2009-03-06 12:28:10)
- RPS 2 OK -> 1 hour
- RPS 3 6 days (The oldest order paid and not deployed is dated2009-03-08 22:27:06)
- RPS 4 8 days (The oldest order paid and not deployed is dated2009-03-06 14:53:05)
- RPS 5 5 days (The oldest order paid and not deployed is dated2009-03-09 15:15:08)

Servers to experiment and to play with:
- Kimsufi L OK -> 1 hour
- Kimsufi XL 4 days (The oldest order paid and not deployed is dated2009-03-10 22:35:11)
- Kimsufi 2XL OK -> 1 hour
- Kimsufi 3XL OK -> 1 hour
- Kimsufi 4XL OK -> 1 hour
- Kimsufi 8XL OK -> 1 hour

Pro, business and critical servers:
- miniSP 2 days (The oldest order paid and not deployed is dated2009-03-12 20:00:29)
- 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
- SP Storage OK -> 1 hour

- EG Bestof 2 days (The oldest order paid and not deployed is dated2009-03-12 20:03:15)
- EG AMD OK -> 1 hour
- EG Max 2 days (The oldest order paid and not deployed is dated2009-03-12 18:20:16)
- EG Large OK -> 1 hour
- EG SSD OK -> 1 hour


- MG Bestof OK -> 1 hour
- MG AMD 10 days (The oldest order paid and not deployed is dated2009-03-05 23:15:08)
- MG Max OK -> 1 hour
- MG Large OK -> 1 hour
- MG SSD OK -> 1 hour
- MG X Large OK -> 1 hour

- HG 2009 BestOf OK -> 1 hour
- HG 2009 BiTurbo OK -> 1 hour
- HG 2009 SSD 48H (The oldest order paid and not deployed is dated2009-03-13 14:11:05)
- HG 2009 Large OK -> 1 hour
- HG 2009 Personalized OK -> 24 hours

Since the changes on the RPS range, we are struggling a bit to follow the rate of your orders. Since the product is stable and has many advantages, we record up to 200 orders per day (not every day, but still). It is also due to set up fee on Kimsufi. We are thus working on the improvement of delivery period. To sum up, it's to see the number of orders for 10/14 days and to organize the production with the number of RPS that you order. It is thus necessary to give us some time to see what we have to produce temporarily or permanently. We also have some bugs left to fix in procedures of RPS up/downgrades. Indeed, very few customers up/downgrade RPS (that proves that Cloud Computing it a virtual concept) and we find some bugs in 1 changes out of 10 that we will fix shortly. In any case, you seem to really like Atom < £10 and 4 cores for < £40! :-)

For Kimsufi, since the installation of the new XL, it is a madness on Kimsufi XL. We are doing our best and I think that next week we will manage to stabilize the production. We still have to fix the delivery of Kimsufi XL with Windows Pro. That will be done this week.

You order a lot of EG BestOf and Max. For the moment we don't manage to produce these servers in sufficient quantities a day. With the starting of the 4th line of server production, this problem should disappear. The installation of this 4th line was decided 4 weeks ago and we are still looking for new team members to reinforce our teams of servers manufacturing. We nicely say that the French job center cause us many problems with our job ads. The training is long and the investment for this training is very heavy. We thus seek for “stable/permanent” team members and not “temporary”. Then, our job ads are filtered by the French job center, which pre-filter applicants that they send to us and at the end we think that it is full employment in Roubaix-City… we thus have started to communicate on the recruitings which we realize to have applicants...It is sad but we are seeking people and we are struggling to advert to receive applications through networks which are there for that. Find out more: http://www.ovh.com/jobs.

We have a technical problem on MG AMD with an incompatibility between the graphic card and the distributions. This problem should be fixed this week and to deliver all delayed servers.

We will increase HG stocks in order to have at least 10 servers of each reference in advance. This enables to to start large infrastructures in a few minutes.

As already announced, after the IP LoadBalancing, we are working on VLAN in “not business" and "business" version. In the same way, the VPN and the firewall. We still have problems on the KVM which doesn't satisfy us. We are wondering how our competitors can propose the KVM with so many bugs and what the customer must say when the KVM is cut when we reboot the server or when the configuration of the KVM is erased. We do not understand. Thus for the moment, we list the bugs to the manufacturers so that they fix them. If it were opensource that would be already corrected… KVM we continue the tests and fixing the bugs.

The equipment for the network Frankfurt/Brussels/Amsterdam/Roubaix left USA and arrive this week (weekend) in Amsterdam. We will deploy everything this weekend and beginning of the week. We should have all working in 10 days (yes!). The equipment for the network Roubaix/London arrives 10 days after and we deploy it also. We connect new 10G transit with Teleglobe and Global Crossing, we rebalance the network and we push 200/220Gbps at the beginning of April. (yes!). Then we add the new routers with… for… (yes!) ;-)

The peering configuration with Orange and TPSA are programmed for Tuesday morning. We should start just after, but the peering with TPSA in Frankfurt will be completely used once the network until Frankfurt is finished.

A new VSS router arrived and will be set up in a few days. Indeed, the current one will be filled in 2-3 weeks. The bug in the IOS of our VSS has to be fixed then the routing is good, which will enable us to look for the network in PT, SE, NO, DK… There is still a lot to do :-)

Regards,

Octave