Performance – Shared hosting
2012-06-11 10:24 At 8:40 this morning a disk synchronization finished running. This in combination with a few processes that had problems caused a performance degradation on and off during last week. Many have felt that it is not as fast as it should. The processes were initially only related to our backup processes but we later found others that also were improperly hammering our disks further degrading performance.
To improve IO performance on this node we are yet again adding more disks to make sure there is room to grow and give great performance at all times. By this afternoon those disks will have started to enhance performance.
Thank you for your patience during this weeks degradation. We will add a number of monitoring points to make sure we do not run into this issue the same way again.
During Friday we solved a few issues with our storage nodes used by our shared hosting environment. As we got performance back to normal on our storage nodes we thought we had the full problem solved. However during the weekend we have seen continued signs of degraded performance.
As we run a large cluster of servers which websites are spread across – we are currently looking at a couple of faulty web servers that are causing certain issues due to errors.
For a period of time we will disable our statistics as we have made updates to our stats programs and want to make sure we did not add something that is not compatible. It will enabled again during the day tomorrow Monday.
If you are one of those customers that has been affected by this we want to assure you we are continuing work on the problem. We will give more updates here as we make progress.