Server NC023 went down for maintenance at 19:22 UTC and was back online at 19:32 after successful completion of its maintenance.
Thanks for your patience.
Server NC023 went down for maintenance at 19:22 UTC and was back online at 19:32 after successful completion of its maintenance.
Thanks for your patience.
We are ready to begin the process of migrating accounts from server NC018 to NC026. We will start with websites on Friday, 3 May at 19:00 UTC.
Once the websites have been migrated, we anticipate that the reduced load on server NC018 will permit re-enabling anti-spam measures. This, I’m sure, will be a welcome relief for all.
We will be performing maintenance on server NC023 (commonly known as the relay server) on Saturday, 4 May at 19:00 UTC. This maintenance will take approximately 10-15 minutes, during which time the server will not be accessible.
We have spent this weekend configuring and testing the new server (NC026). Our next step is to conduct a test migration to determine how long it will take and what possible issues may arise that we’ll need to address.
When we have done this, we will formulate a migration plan, determine a window in which the migration will take place, and communicate that information to you.
Once that’s done, we’re looking forward to having a much newer, much faster, more resilient, more secure, and more configurable server. I know you are too.
On Thursday the 18th we set up a new server in a new data centre. We’re excited about the innovations in security that this new configuration will bring to your hosting account.
We are working diligently to configure and test this server, and test the migration process to minimise the time it will take. Due to issues on the old server, our priority is to get the migration done as soon as possible. While we had hoped to do this over this weekend, we may need to do it during the week. Either way we will ensure that any disruption will be kept to a minimum.
We will post further updates here over the course of the next few days.
Contrary to previous information, the replacement server will not be ready until mid to late next week.
The good mews is that we have taken further steps to mitigate the performance issues, and we have seen no evidence of performance problems in the last 48 hours.
We’ll post another update here when we have firmer time lines to tell you about.
Server load today has been within tolerance, and email has not been backed up for more than about a couple of minutes at any point.
We’ll post an update later today on the acquisition of new hardware.
A contributory cause of the performance issues on server NC018 is one of the anti-spam systems on that server. Unfortunately, we have had to deactivate part of the anti-spam system, and we’re already seeing the results in the form of spam that would normally be blocked reaching mail boxes. We apologise for this temporary issue. On the plus side, the anti-spam systems on the new server will be an improvement over the one on the old server.
As we informed you on 6 March, we had intended to increase the memory (RAM) on server NC018 on 9 March. (Server NC018 hosts the majority of websites, and all email except specialised email services.) This maintenance would have solved performance issues that were happening with more regularity. However, the maintenance failed due to hardware compatibility issues.
When the maintenance failed, we immediately began preparing to acquire new hardware to replace this server. During the week of 11-15 March we obtained a number of proposals and advice on a plan we had begun developing some months ago.
However, on 18 March the mail server was overwhelmed. The most obvious result of this was that email delivery was delayed during peak periods by as many as several hours. While it was not clear exactly why the mail server was overwhelmed, there was an obvious way to deal with the problem. We implemented this, and the mail queue was cleared in a short amount of time. We believed that this approach should have worked to prevent the mail server from being overwhelmed if the mail server again experienced heavy volume, but it did not when the same thing happened on 19 March.
We are therefore urgently preparing to set up new hardware. I don’t have an exact time line for you at the moment, but it is our goal to have the hardware in place and online by the weekend, and we will endeavour to transfer everything as quickly as we can over the weekend.
We will update you on major events by email, but please check our status page once or twice a day at status.niner.net or status.ninernet.net for progress reports.
We apologise for the inconvenience that this transition is causing, and ask for your patience as we go through this upgrade.
We have cleared the mail queue on server NC018, and the web server has been brought back online.
Systems at a Glance:
Loc. | System | Status | Ping |
---|---|---|---|
NC023 | Internal | Up? | |
NC028 | Internal | Up? | |
NC031 | Internal | Up? | |
NC033 | Operational | Up? | |
NC034 | Internal | Up? | |
NC035 | Operational | Up? | |
NC036 | Operational | Up? | |
NC040 | Internal | Up? | |
NC041 | Operational | Up? | |
NC042 | Operational | Up? |
Subscriptions:
Search:
Recent Posts:
Archives:
Categories:
Links
Tags:
Resources:
On NinerNet: