C-Servers | The Pub

cservers

Thread to bring out some news regarding C-Servers, general chit-chat, and some good talks on the rocks. Some good campaigns and coupons may land here as well.

web.c-servers.co.uk

yoursunny

We don’t buy C-Servers unless it’s the 99-year plan.

kf5ydu

I think it’s good for users to have control over selecting what resources they need. So long as it actually provides a discount and isn’t used to raise prices of existing resources like a lot of large companies would do.

As long as a company is transparent on their pricing and comparisons I think it’s an excellent solution. That’s one of the things I love about this forum, we can actually talk to the people selling us things.

cservers

Nothing that the FULL15 coupon is still working and will continue to work until the end of the month; and that new coupon codes and percentages will be provided here shortly.

We also continue improving geolocation and looking forward to improve peerings to specific peers on our ASN.

cservers

We are aware of a networking fault at Delta1 which is causing the server to be inaccessible since 06AM and have escalated to the correspondent agent in order to sort the issue out as soon as possible.

There seems to be no hardware fault as consumption has continued during this period.

EDIT: Seems the counter is not correct. Power maintenance at Telehouse until 6PM. We do apologize. https://status.23m.com/maintenance-windows/63671

edmarhub

@cservers said:
We are aware of a networking fault at Delta1 which is causing the server to be inaccessible since 06AM and have escalated to the correspondent agent in order to sort the issue out as soon as possible.

There seems to be no hardware fault as consumption has continued during this period.

EDIT: Seems the counter is not correct. Power maintenance at Telehouse until 6PM. We do apologize. https://status.23m.com/maintenance-windows/63671

@cservers can you check your Amsterdam Full VPS I think it is down for a couple of hours already, also the ai support is bugging can't proceed the login!

cservers

@edmarhub said:

@cservers said:
We are aware of a networking fault at Delta1 which is causing the server to be inaccessible since 06AM and have escalated to the correspondent agent in order to sort the issue out as soon as possible.

There seems to be no hardware fault as consumption has continued during this period.

EDIT: Seems the counter is not correct. Power maintenance at Telehouse until 6PM. We do apologize. https://status.23m.com/maintenance-windows/63671

@cservers can you check your Amsterdam Full VPS I think it is down for a couple of hours already, also the ai support is bugging can't proceed the login!

We do have a hardware fault at Amsterdam. It was already escalated to the upstream, which acknowledged the message we've sent and will be checking in order to sort it out. Over the past week things were progressively less stable due to one of the RAM ECC sticks there posting more errors than usual (without corruption risk - all were corrected) and probably that module has now packed up for good - will need a replacement. That had been detected already on September 20th and a ticket had already been proactively opened.

Peerings may be slower as well at AS214566, temporarily, for the same reason.

We do apologize for the inconvenience. All updates first-hand at https://portal.cfld.uk/status.

Scioner

@cservers said:
We do have a hardware fault at Amsterdam. It was already escalated to the upstream, which acknowledged the message we've sent and will be checking in order to sort it out. Over the past week things were progressively less stable due to one of the RAM ECC sticks there posting more errors than usual (without corruption risk - all were corrected) and probably that module has now packed up for good - will need a replacement. That had been detected already on September 20th and a ticket had already been proactively opened.

Peerings may be slower as well at AS214566, temporarily, for the same reason.

We do apologize for the inconvenience. All updates first-hand at https://portal.cfld.uk/status.

Btw, AI chat bot escalated my issue to the ticket the moment it checks it can't access the server, even though I didn't ask for escalation. I'm pretty sure there was already a pile of tickets like that, as I waited relatively long before even talking to the bot.

Probably there should be something like "bot database", which records recent issues. This way bot can answer with "we had reports about that, thank you".

cservers

@Scioner said:

@cservers said:
We do have a hardware fault at Amsterdam. It was already escalated to the upstream, which acknowledged the message we've sent and will be checking in order to sort it out. Over the past week things were progressively less stable due to one of the RAM ECC sticks there posting more errors than usual (without corruption risk - all were corrected) and probably that module has now packed up for good - will need a replacement. That had been detected already on September 20th and a ticket had already been proactively opened.

Peerings may be slower as well at AS214566, temporarily, for the same reason.

We do apologize for the inconvenience. All updates first-hand at https://portal.cfld.uk/status.

Btw, AI chat bot escalated my issue to the ticket the moment it checks it can't access the server, even though I didn't ask for escalation. I'm pretty sure there was already a pile of tickets like that, as I waited relatively long before even talking to the bot.

Probably there should be something like "bot database", which records recent issues. This way bot can answer with "we had reports about that, thank you".

That is actually an excellent idea - the agent can search the topics of existing tickets and if there is a majority of the same topic it can simply state that and notify the systems department

There was indeed a certain ticket peak at that time. Per usual whenever a fault like these happens... at all providers.

We think it's somewhat inefficient having all tickets talking about the same... only adds pressure to something the provider is already aware of. And time spent answering the tickets can be better spent pushing ahead to fix the issue as soon as possible. Probably a good use of automation there.

Wonder_Woman

Hey @cservers, I hope you're doing well!

I just went to check on the status of a ticket that the chat said it had opened for me almost a week ago, and I was surprised to see there's no ticket! I'm trying to talk with the chat bot again now, and all I'm getting is:

"Sorry, I'm having some issues. Can you try asking your question again?"

Repeated, no matter what I type.

cservers

@Wonder_Woman said:
Hey @cservers, I hope you're doing well!

I just went to check on the status of a ticket that the chat said it had opened for me almost a week ago, and I was surprised to see there's no ticket! I'm trying to talk with the chat bot again now, and all I'm getting is:

"Sorry, I'm having some issues. Can you try asking your question again?"

Repeated, no matter what I type.

Hello! We've been getting tickets yesterday and today - this could be sometimes a temporary unavailability from the underlying AI platform but should clear out on itself. We've just tested and it's bringing answers as of right now.

edmarhub

@cservers said:

@edmarhub said:

@cservers said:
We are aware of a networking fault at Delta1 which is causing the server to be inaccessible since 06AM and have escalated to the correspondent agent in order to sort the issue out as soon as possible.

There seems to be no hardware fault as consumption has continued during this period.

EDIT: Seems the counter is not correct. Power maintenance at Telehouse until 6PM. We do apologize. https://status.23m.com/maintenance-windows/63671

@cservers can you check your Amsterdam Full VPS I think it is down for a couple of hours already, also the ai support is bugging can't proceed the login!

We do have a hardware fault at Amsterdam. It was already escalated to the upstream, which acknowledged the message we've sent and will be checking in order to sort it out. Over the past week things were progressively less stable due to one of the RAM ECC sticks there posting more errors than usual (without corruption risk - all were corrected) and probably that module has now packed up for good - will need a replacement. That had been detected already on September 20th and a ticket had already been proactively opened.

Peerings may be slower as well at AS214566, temporarily, for the same reason.

We do apologize for the inconvenience. All updates first-hand at https://portal.cfld.uk/status.

Any ETA? Seems like it is 3days and 6hrs already since the last time I got a probe response from the server? Any plan compensation like extended expiration while the maintenance is on progress?

cservers

@edmarhub said:

@cservers said:

@edmarhub said:

@cservers said:
We are aware of a networking fault at Delta1 which is causing the server to be inaccessible since 06AM and have escalated to the correspondent agent in order to sort the issue out as soon as possible.

There seems to be no hardware fault as consumption has continued during this period.

EDIT: Seems the counter is not correct. Power maintenance at Telehouse until 6PM. We do apologize. https://status.23m.com/maintenance-windows/63671

@cservers can you check your Amsterdam Full VPS I think it is down for a couple of hours already, also the ai support is bugging can't proceed the login!

We do have a hardware fault at Amsterdam. It was already escalated to the upstream, which acknowledged the message we've sent and will be checking in order to sort it out. Over the past week things were progressively less stable due to one of the RAM ECC sticks there posting more errors than usual (without corruption risk - all were corrected) and probably that module has now packed up for good - will need a replacement. That had been detected already on September 20th and a ticket had already been proactively opened.

Peerings may be slower as well at AS214566, temporarily, for the same reason.

We do apologize for the inconvenience. All updates first-hand at https://portal.cfld.uk/status.

Any ETA? Seems like it is 3days and 6hrs already since the last time I got a probe response from the server? Any plan compensation like extended expiration while the maintenance is on progress?

Yes, extended expiration is a possibility, additional days are a possibility, and so it goes.

Pinged the upstream for this one. We want this up and working as soon as possible.

cservers

Alpha1 Amsterdam back online, but with two caveats I'm mentioning here:

• the session with AS214481 is temporarily down, so we're rerouting with Hop179 also in AMS as a fallback temporarily, and it's all temporarily limited to 1 Gbps. We could have prioritized Coventry for that, but it would have added latency and we preferred location over speed, given that the plans were announced already with 2 Gbps maximum.

• the server is not yet stable, as the defective RAM module - or the server change - was not dealt with over the weekend, due to being the weekend (as we suspected and mentioned beforehand on the Status portal). Therefore, we have tested the defective RAM module, and found that selectively limiting the RAM blocks used to around 50% of the capacity lowers errors to at least a level where VM downtimes don't (or shouldn't) occur. But it's not okay - it's not a long-term viable procedure nor is it a replacement for any change. Just a stop-gap solution to avoid having the VMs offline and falling periodically. No data is at risk, the errors are all corrected as well, but we're watching this one.

EDIT: nope, it's not right, just locked up again; we're preventively stopping all VMs at Alpha1 for security reasons.

cservers

Okay, this time hopefully for good stability-wise: after a long journey and hard work, Alpha1 Amsterdam is back online with all VMs, and all pending orders for Alpha1 were fulfilled.

The caveat for AS214481 continues to apply, so we're rereouting to Hop179 in the meantime as a fallback, limiting things to 1 Gbps.

The server, however, is now stabilized temporarily, with several management actions taken on our end:

• we're putting all VM loads on the remaining working RAM, forcing them to never go to the defective module, deliberately;
• several RAM contention management measures were applied on our end to ensure this would work, the most visible of which will be a temporary and one-off 20% balooning, to ensure contention/shared memory and some added stability/uptime, for all users with 936 MiB / 1GB RAM or more - those with the 512MB plans already have little of it, we're leaving them at 0;
• ZFS ARC (cache for the internal filesystem we use) was limited to the minimum possible, as it can't be contained to exclude a single RAM module, and this will mean disk speeds will be lower for now in tradeoff for the natural stability that brings on RAM access;
• tests were done pre and post-raise of the VMs, and naturally because we're not using the defective module, we're seeing 80x less errors in runtime, and a satisfatory level of errors at present which ensures peace of mind for everyone.

We're also suspending temporarily the sales for 6GB and 8GB RAM plans for Alpha1.

We'll continue to work with the upstream to fix both technical faults occurring on their end. As soon as we are able to ensure a normal service level again, things will get back to normal on all fronts.

cservers

RAM replaced at Alpha1 with zero errors, server 100% working.

As soon as we get the connectivity with AS214481 up and running again speeds will improve in accordance.

2423750524

Hello, when will the server speed of Alpha1 Amsterdam be restored? The speed to various locations seems to be a bit slow now

cservers

@2423750524 said:
Hello, when will the server speed of Alpha1 Amsterdam be restored? The speed to various locations seems to be a bit slow now

Hey! IPv4 had already been restored with 10 Gbps at the ASN - we have been waiting on the upstream for the IPv6 session to go accordingly restored as well, and that one's still at 1 Gbps for now.

cservers

Support over AI will get a major overhaul and improvement, with a different model: AI starts to be the intermediary, not the final destination, and will start to be humanly supervised on the support channels.

Also, all Live Chat conversations will start to have a reference assigned, just like they were a ticket, for the sub-12 USD customers.

More on that to be divulged later today (08/10).

For the same reason, you may experience today several periods of unavailability and instability on the Live Chat and the ticketing section, while we implement these changes.

This is one of the few relevant occasional painpoints with our services that are remaining - that ends today.

mioakash

Hi @cservers,
My FullVPS-1024 Essential (#6849) is Active, but I can't start the VPS or install an OS. Both actions return EK errors.
The support bot acknowledges it's an infrastructure issue but won't open a ticket, saying I'm not eligible for human support.
I paid €10 with a coupon, but the regular annual price was over €12. I saw your earlier post confirming that FullVPS-1024 should be eligible for support.
Could you please look into this and help get my VPS working?
Thanks!

kf5ydu

My renewal is only applying the top40 promo code I initially used towards half of my order. Can you please assist with this?

kf5ydu

@kf5ydu said:
My renewal is only applying the top40 promo code I initially used towards half of my order. Can you please assist with this?

If it helps here is the original invoice.