Community Tip - Need to share some code when posting a question or reply? Make sure to use the "Insert code sample" menu option. Learn more! X
Hi,
We have two replica servers in one of our environment. with one of the replica we are not finding significant difference in performence in accesing the object compared to accessing the same with MASTER as prefered file server.
So I want have following two concerns regarding that ..
1) what are the way by which we can ensure the object is actually fecthed from indented replica server.
2) and my secodn question is that, when I see MS log of one replica , it is trying to download the part from another replica server. ( Download URL has name of the another replica server ) ..
I do not under stand what configuration I have missed .. please request to provide some suggestion and opinions from you end ...
Hi,
Normally, the reason a file is not downloaded from Replica A (presumably set in your user preferences?) is because the exact iteration of the file does not exist on Replica A and hence it has to be pulled from elsewhere.
For example, someone at a site with a different Replica B, or someone at the Master site, may have checked in a later iteration of the file to what is on Replica A. This latest file then might only exist on the Replica B (if it has not been Sync'd back to the Master) or on both the Master and Replica B (if it has been Sync'd). It is not until the next Replication schedule that includes a rule for this file, or until someone with Replica A set downloads the file that it actually gets transferred to Replica A for later use.
There are two distinct operations for Replica's -
Hence, I presume that you have created and run a replication rule & schedule to replicate the file you are testing to the Replica you are testing the download from?
It's possible to set a Replica's proximity to other replica's so that data for one replica is pulled from another Replica that is closer to it rather than the Master. ie you could have a Master site that Replicates to Replica A and then there is a further hop away from the Replica A to a Replica B. The data has to have Replication Rules & Schedules to ensure that data that is needed to be replicated to Replica B is available to replicate from Replica A, otherwise it will have to come from the Master site.
Rgds
Gary
I just re-read your post...
If I wanted to be sure of where the data was coming from on the client machine in my testing, I would install Wireshark on the client machine, setup a couple of hosts in the filter line for the master and each of the replica server's hostnames (so that I only captured relevant traffic) and then watch the traffic patterns when I did the download of the file.
Rgds
Gary
Manoj,
Did the information in Gary's posts answer all of your questions about your replica's behavior? Let us know if you still have any questions, or are seeing any other issues with your replica site.
Yes, it was helpful .. I am yet to try with wireshark tool ...
thanks to Mr. Gary Mansell
I will post the details once I complete my test with wireshark tool ..