Knowledge Base Article
Was this article helpful?
Greetings, doylc!
Having your journal index at the Brief level only means you can't search for phrases. Any other indexable and searchable attribute is still available to you. For example, with Brief level, you can't search for phrases such as "Discovery Accelerator Administrators Guide", but you could search for the individual words. Having those individual words on their own lines would have your search criteria look something like this:
Discovery OR Accelerator OR Administrators OR Guide
If you put all of those words on the same line, the logic changes to something like this:
(Discovery AND Accelerator) OR (Accelerator AND Administrators) OR (Accelerator AND Guide)
Of course, the actual logic depends on the selection of ALL OF or ANY OF for the criteria field, so the above examples may not be what you actually get. To see what your search criteria really looks like when passed to the search engine, obtain a dtrace of the EVIndexQueryServer process on the EV Indexing Server while starting a DA search. Look in the dtrace log for some part of your criteria (in the above example, you could look for any of the 4 words). The line found would contain the logic of the criteria passed.
As for the disk space on the EV Storage Service server and our recommendation for 40 GB minimum of free space on the drive containing the Vault Service Account's TEMP folder, this recommendation is based on the heavy usage of this TEMP folder during normal archiving operations as well as exporting. This TEMP folder is where EV puts its temporary files while converting new items to HTML or text prior to archiving, and also when converting from our DVS files into MSG files. You'll get better performance from your system if the VSA TEMP folder is located off of the system drive (i.e., C:\) and off of a drive that hosts a pagefile.sys file.
Finally, EV 10 has the great feature of being able to search indexes while they are being rebuilt. Based on your system's configuration, you can rebuild any index volume or entire archive's index and still search the 'old' index volumes until such time as the 'new' index is ready and the system switches to the 'new' index volumes. I recommend you test this in your environment, though. Create an archve for test user's mailbox and archive some items into it. Then run a search against that archive to be sure you can find items. Once you've confirmed you can find items, rebuild that archive's index while running another search. You should be able to find the same items in the original index volume while the new index volume is being prepared and populated.