Can anyone provide any pointers why the charts always hide the spikes. For example I’m charting CPU usage on one of our servers where usage is usually at about 5%. I get it up to 100% and that is reflected in the live chart, but after a day or two I won’t see that spike at all.
Same thing for bandwidth…
Be careful with raw keep time, if your database goes over 2gb in size the dude will crash.
Thank you both, I’ll try to keep it under control. Any recommendations, I only have about 50 devices. Also where can I check the database size, can I shrink it afterwards when I’m approaching the 2GB limit?
Oh go crazy then
you are not probably not close to 2gb, just for reference I was unable to have 2 weeks of raw keep time but I have 300 devices and probably 450 graphs ~1200 probes. Search for vacuum in this forum. The database is easy to find in the dude folder. I vacuumed mine once and it went down by a bunch ~1.8 to 850ish. There are already a bunch of good posts about what is in the database. No one has taken it very far though. The data is compressed inside the tables so no one has done much other than vacuum or try to repair a broken record. No tools exist to extract a single graph or find/remove corrupt data or remove a single value that is erroneous from a table.
Although there is some code on this forum that will extract data from the web pages that the dude generates… I seem to remember something like that anyhow 
GL
Lebowski