HISTOGRAM information from log files

After a 100K test has finished, the log files are located in /cdf/scratch/cdfprod0/4.8.0pre3_05/test1/logs/, for example.
To make histograms summarizing information from the log files:
- First, make text files with just the information of interest. The scripts take two inputs, the name of the directory in cdfprod0 to indicate the release, and the output text file, named appropriately:
```
     cd /cdf/scratch/cdfopr/mon
     source .source_me
     make_ntuples.sh /cdf/scratch/cdfprod0/4.8.3_07/Stream_A 4.8.3_07_Stream_A 
```
  for the experts: if make_ntuples.sh is missing, do
```
     make OfflineMon.bin
```
- run root and plot the distributions for memory consumption and db access statistics:
```
        root.exe
        .x plot_db_stat.C("4.8.3_07_Stream_A.db.root")
        .x plot_mem_stat.C("4.8.3_07_Stream_A.mem.root")
```
- put plots into the offline elog

SUMMARIZE the errors and notify the experts

To summarize the error messages from the 100K test,

     cd /cdf/scratch/cdfopr/mon 
     ./make_error_summary.sh 4.8.0pre3_05 prod-errors-480pre3.txt

The output text file lists each different error that occured during the 100K test, the number of times that it occured, and the file name, run number and event number of one of the occurances.
Click here for list of errors which can be ignored according to the relevant experts. In general, these errors should be reported if they begin to occur much more frequently than usual (e.g. 300 occurrences in 100K test increases to 3000 occurrences in 100K test).
For each error:
- See list of librarians to find the librarians responsible (or ask Pasha)
- Send the appropriate librarian an e-mail telling them what error occured, how many times it occured, and the file, run and event number of the event listed in the error summary.