Job Completion % bug
This has been an issue inn 11.X as well, not sure why it was ported to 12.X
Say you have a job with 2 archive locations (the primary location and the secondary it needs to replicate to).
This how it works (albeit with more intermediate steps):
1 Job starts 0% job completion
2 Archive in the first location is starting to get saved 18%-25% job completion
3 Archive in the first location is finished saving 95%-98% job completion
4 Archive starts the transfer from the first to the second location 98% job completion
5 Management server reports job as "frozen" on large archives
6 Archive is done replicating 100% job completion "frozen" status is cleared
This percentage measuring system is wrong, because even when the secondary location is on the same network it will take time for the archive to get transferred, if the secondary location is on WAN, which is how it should be to make any sense, the transfer will take longer than the backup itself.
Now, if this was just a "cosmetic" issue, then who cares, right? Well, 12.5 now has a wonderfully "smart" self monitoring that detects when the jobs are "frozen" and throws up an alert. Try backing up a terabyte with the secondary location being connected via internet, and will see the "job frozen" error. As far as the management server can tell the job sits at 98% for several hours, all because the secondary storage node file transfer progress is ignored. I mean how hard could it be to feed some progress info back to the management server? You already have the archive size available from the first location. If the system showed 50% completed after creation of the first archive, and updated the second storage node progress, then it won't appear "frozen" to the monitoring.
It would be great if the developers looked at this!
11.7 used to do the same thing, but without errors, it would simply dump the job from the main management interface and let it silently continue inside the direct connection to the agent UI.
thank you,
tony
