It has been fixed already, it should be slowly recovering.

As a reminder, new replicas are more likely to be up to date than old ones (see they have 0 lag)- this is not on purpose- but newer hardware and extra redundancy make easier to recover from problems coming from production like this one. So please start using the new hosts ASAP, they are both faster and more reliable.

In T177505#3661118, @Marostegui wrote:

It has been fixed already, it should be slowly recovering.

Cool!

In T177505#3661148, @jcrespo wrote:

As a reminder, new replicas are more likely to be up to date than old ones (see they have 0 lag)- this is not on purpose- but newer hardware and extra redundancy make easier to recover from problems coming from production like this one. So please start using the new hosts ASAP, they are both faster and more reliable.

XTools actually is. I simply opened this task because I saw lag on the old systems. We're looking at removing our replag notification from the web interface soon.

Thanks!

Replag is now at 0 according to https://tools.wmflabs.org/replag/