Hi,
We have been investigating the issue that caused a number of disk servers to loose their network connectivity yesterday (Thursday 12th April). There was one incidence of this early yesterday evening, and none since then.
From an analysis of the problem we can see this has been confined to Atlas disk servers (in fact AtlasDataDisk so far) in two particular batches of disk servers.
Yesterday afternoon updated kernels were applied to four disk servers and these have run successfully overnight. The same update is now being applied to the remaining disk servers of the affected batches that are in AtlasDataDisk.
There is a 'warning' declared in the GOC DB for this problem. We will let this expire at midday local time (11:00 UTC).
Gareth Smith.
|