WE have the same problem,
Only real thing you can do is quatlity of service over that line.
If not.. well update those computers manually or upgrade your line
So everytime we have huge updates/upgrades the line take some punishment.
even we have those reposerties on that side of the line. the update still need to go over that line still.
wich make alot of our collegues unhappy.
1 of 1 people found this helpful
Hmmm, I know with EPO 3.6.x (I assume 4.5 has this as well) you can just deploy the delta DAT's every day and then on the weekend push out the bigger ones.
We have lots of WAN links that are just T1 and the daily DAT's take a couple of minutes for each site (granted, we deploy via NCP to Novell NetWare servers, but I imagine SMB to Windows would be just as fast). HTTP tends to be a little chatty with all the GET and responses.
I don't believe that pushing to distributed repositories has any sort of retry or failback methods. basically I think you schedule it to run once/day and go from there.
We also rsync out the data at night just in case.
Thanks for the help so far.
I've managed to halve the amount of data by removing the V1 dats since I can get away with only V2 (I'm using Agent 4.5, VScan + Antispyware 8.7 only). I'm trying out another method at the moment which is a bit of a bodge but should do it - Robocopy the contents of the <repo>/current/VSCANDAT1000/dat/0000 since this supports resuming, then run a replication (which won't do anything as it's already up to date). Luckily this won't have to be done every day, every 3 days or so is fine for me.