We have two 570 systems, which have a dedicated 1 Gb Ethernet link between the machines for replicating "Realtime" data as part of their HA solution. The replication software has a lag threshold setting currently set to 20 mins and warns when this is reached. Initially there was 1 queue setup to send the data between the systems and this has been increased to two and more recently to three to keep it under 20 mins to try and resolve this but with limited success.
During the last two weeks however we are starting to see the threshold breached each day with one queue becoming an hour and a half behind in lag time. I'm trying to ascertain if the link is becoming 'saturated' , which I will need to raise as a risk to the clients business if this is the case.
Currently there are three send jobs from source to target.
Anyone ever experienced the 'lag' effect and how to circumvent it?
October 28, 2008 3:48 PM
October 30, 2008 7:17 PM