This error means that the chunk was saved to some sig block but can not be read from it. It could be because something happened to the cache, for example sig blocks were overwritten (FIFO). If the aggregation errors are few or occasional, these errors should not impact performance. Regards Pete.. ... View more

Rafal, Let me see if I can comment by number here. First off, all of your comments on #1 are correct with the exception of mentioning the root CE. In straight transparent caching where you have a router and CE at a branch location running wccp, you do not need a root ce. This would be simply web caching to the internet. In #2, this is all correct. Root CE would be at the home location along with a CDM for CE and channel mgt. You would probably want to put another CE in the same location as the root so it will become the root CE in case of a failure. The only thing that is missing in this discussion is the use of a Content Router. This is where you have a coverage zone file on a Content Router. Basically what happens is a client requests content. All requests hit the Content Router first and then based on the coverage zone file, the client is then sent to the correct CE for content. This is the most complex use of ACNS. #3 is correct. In #5 you do not need wccp all the time when using preposition content. You could have a senario where a client his the CE in a proxy configuration where no wccp is used and grab prepostioned content. You are correct in the sense that wccp is simply an intercept senario which then sends the port 80 or 8080 type request to a content engine. ... View more

Dom, Yes, that is a bit confusing. Let me try to explain. So my assumtion now is you are running something like this: 3.0(0)A1(6.2) For the purpose of our conversation, this is considered ACE A1. When someone mentioned 2.0, the code will now be A2 as opposed to the A1 above. This is what 2.0 means. I am not completely sure as it's not released yet, but the first rev of 2.0 ACE would be 3.0(0)A2(1.0) (note the A2) Regards Pete.. ... View more

Hi Joao, Unfortunately, the SCM's in the CSS are redundant only from a hardware perspective. They cannot be configured to take over from a "software" perspective when an interface goes down. I guess one potential thing to do would be to have a customer script with a keepalive to test ip connectivity and when/if that keepalive in software fails, have the script run the commands to take the SCM down. Regards Pete.. ... View more