Webdock - Notice history

On monday we will be tidying up wiring in a rack in our Denmark DC for better airflow and accessibility. The procedure involves replacing fiber uplinks one by one. As we have a fully redundant and routed infrastructure, if no mistakes are made, then our customers should not notice anything happening during this maintenance window. As there is always a small chance that a mistake happens or a cable gets snagged in which case a brief network outage could happen. For this reason we are announcing this maintenance window.

This maintenance has been rescheduled for Thursday January 30th, due to some required equipment not making it to us in time.

On Thursday we will be performing some electrics related maintenance for the rack which tripped fuses when we did our generator upgrade and power outage tests the other day. The maintenance consists of putting less sensitive fuses on both of our circuits and potentially replacing a power transfer switch. We hope that the work will happen without incident, but there is a chance that power may go away from that rack again - we cannot guarantee this will not happen. This will impact most of our Ryzen workloads, our Epyc workloads and our core services such as our dashboard. If a power incident occurs, we will be operational again within about 5-10 minutes and our entire team will be standing by to make sure everything comes up as quickly as possible after a potential failure.

We absolutely need to solve this issue as the situation is now, then if we have an unplanned power outage there is a chance this rack will go down, and we don't want that. In fact, everything we've designed in our DC is designed to prevent exactly this. We hope to get this issue resolved fully on thursday and hopefully without incident.

On Tuesday January 28th we will be performing some electrics related maintenance for the rack which tripped fuses when we did our generator upgrade and power outage tests the other day. The maintenance consists of putting less sensitive fuses on both of our circuits and potentially replacing a power transfer switch. We hope that the work will happen without incident, but there is a chance that power may go away from that rack again - we cannot guarantee this will not happen. This will impact most of our Ryzen workloads, our Epyc workloads and our core services such as our dashboard. If a power incident occurs, we will be operational again within about 5-10 minutes and our entire team will be standing by to make sure everything comes up as quickly as possible after a potential failure.

We absolutely need to solve this issue as the situation is now, then if we have an unplanned power outage there is a chance this rack will go down, and we don't want that. In fact, everything we've designed in our DC is designed to prevent exactly this. We hope to get this issue resolved fully on tuesday and hopefully without incident.

Today and tomorrow we are doing some general network maintenance. The changes made are to improve our routing within our Denmark DC. The changes should not affect any customer workloads other than a tiny percentage of VPS servers which currently have erroneous configuration. In cases where a VPS is affected, it is only impacted for a few seconds as routing is being reconfigured. The impact should be minimal and likely not noticeable by our customers. If you do experience some sort of network issue that lasts more than a minute or two, please reach out to our support team.

We have a host machine in Canada which needs a kernel upgrade and reboot. Your VPS will go down for at most 5-10 minutes, usually a lot less. You can check if your server(s) are affected by logging in to Webdock as this will then be shown at the top of the page in a red alert window.

This network maintenance has been on-going in stops and starts for the past week or so. We expect the final changes to be completed within the next 2 days. These changes should have NO impact on your server or workloads.

The maintenance will start in a few minutes. We expect your VPS will go down for somewhere between 15 and 30 minutes depending on how things go. Our target is no more than 20 minutes for the entire operation.