philrodo Posted August 7, 2014 Report Share Posted August 7, 2014 I started a new case in Intella yesterday and had it process and index 14 PSTs totaling 43 GB. It took Intella more than 10 hours to process at an average speed of 853 items per minute. Isn't this awfully slow? It also produced some 69K+ exception errors, which I haven't looked at yet. This is on a pretty fast machine, i7 eight-core, 16 GB of RAM (which I Intella never used more than 5-6 GB when I checked it and CPU usage was hardly maxed, although at most times all eight cores were being used by Java). We also used three drives on the machine, one to read the PSTs from, the C:\ volume to temporarily write/read the items to, and another drive to write the Intella case to. All three drives were SATA drives directly connected to the computer. So what will cause this slow processing speed? Is there a way to improve on the processing speed? Link to comment Share on other sites More sharing options...
AdamS Posted August 8, 2014 Report Share Posted August 8, 2014 At a guess I'd say the complexity and break down of the data contained in the PST's probably caused the slower indexing rate. The fact that there are so many exception items also makes me think there could be data corruption or other issues that may slow Intella down. Link to comment Share on other sites More sharing options...
philrodo Posted August 11, 2014 Author Report Share Posted August 11, 2014 Adam: Thanks for the feedback. But I keep on processing more and more PSTs (all from the same server) and I keep on getting the same slow speeds. And frankly, in our experience processing speed with Intella is a hit or miss proposition--sometimes, it's relatively fast, other times it's painfully slow. And sometimes, it keeps on crashing and won't process large number of files, which then have to be broken down into bunches to feed into Intella... Although we're a long way from the early days of Intella where it was painfully slow, processing still remains a hit and miss proposition. Yeah, we can blame it on corrupt data, but that's the data we have to process... Best regards, Phil Link to comment Share on other sites More sharing options...
Recommended Posts