You may be familiar with all the things Giganews does, like increasing our retention, launching encrypted Usenet access, and developing the Giganews Accelerator, but do you know about all the things that Giganews doesn’t do?
A complex server cluster like Giganews’ Usenet system requires maintenance all the time. We have to replace hard drives, swap out servers, update operating systems, and a whole host of other tasks.
Even though from time to time we need to take down certain parts of our cluster, we never actually have to take our service off line. This means Giganews’ service is always available and isn’t subject to maintenance windows similar to what is experienced through other Usenet servers. We are also able to take down elements of the network without negatively affecting the download speeds of our customers.
One of the more common problems people report with Usenet servers are slow download speeds related to increased traffic. This is most often experienced during the period of increased Usenet usage over the weekend, called the “weekend rush”. When news servers have very little capacity, customers experience slow download speeds when usage increases.
Giganews maintains extremely high levels of capacity so that weekend rushes or increased usage driven by service improvements (i.e. 200 days binary retention) don’t affect the speeds at which our customers can download.
Giganews has completed many upgrades to our storage and server clusters over the years. These upgrades include adding and replacing storage, adding servers, and moving to new data center locations. These upgrades are generally seamless for our customers and very rarely even noticed.
For most of Giganews’ customers the only sign that an upgrade has taken place is all the extra retention!
Many Usenet systems, in trying to keep pace with Giganews’ break neck retention upgrades, experience problems almost immediately after an upgrade. These problems include lost articles, downtime, and roll backs (undoing the upgrade).
Giganews is a bit spoiled because we have a top-notch team of engineers and programmers who have developed a custom, in-house Usenet server application. This custom Usenet server application was developed 100% in house and utilizes proprietary code (it’s not an open source or third party Usenet application). Our server software is designed to let us upgrade both code and hardware without interrupting service to our customers.
Many Usenet servers will advertise retention levels which don’t exactly match their actual retention. They may carry long retention in certain groups or even forgo carrying some groups all together in an attempt to reduce the size of the news feed.
These techniques allow certain Usenet servers to advertise “up to X days retention”. While this may be technically correct (even if the advertised retention is only available in one group) the reality of the situation is that these claims aren’t exactly accurate.
Because of Giganews’ robust Usenet system and our vast storage capabilities, we don’t have to hand pick newsgroups in which to offer long retention. We also carry the full suite of newsgroups, so we’re not reducing the number of newsgroups we carry in an attempt to make it seem as if we have more retention than we really do.
Giganews is able to stand by our service levels without having to use “funny” retention statistics to make our service look better than it really is.
We’re always excited to announce improvements to our service, but writing announcements about how everything is okay isn’t exactly exciting news. We don’t talk a lot about how our upgrades, maintenance, and retention management all run smoothly, but we feel this is an important part of the value Giganews provides to our customers.
The next time you notice Giganews increasing retention or adding features to our service, remember that there is a lot of complex work going on behind the scenes to make sure we’re delivering the level of service you expect without having to interrupt your access or slow down our service.