[Bioclusters] speaking of grid engine 6.0 ... I've made a "5.3 vs 6.0 features" listing

Chris Dagdigian bioclusters@bioinformatics.org
Thu, 06 May 2004 16:11:51 -0400



Bernard Li wrote:

 > Hey Chris:
 >
 > Nice document!
 >
 > My understanding of the 'db' enhancements of SGE 6 is:
 >
 > Spooling information is stored in berkeley db
 > Accounting information is stored in SQL
 >
 > So even if you use the 'open source' version of SGE6, the accounting
 > information should still be stored in the SQL db but the modules to
 > extract the information is provided by the product from Sun?
 >
 > In that case at least we could have some third-parties create some
 > account reporting tools based on XML/SQL queries.
 >
 > Thanks,
 >

hmm...

I have the open source grid engine 6.0beta2 built from CVS source on my 
powerbook and there is no mention at all about SQL storage of accounting 
data.

The 'man accounting' command still shows the old info. The manpages are 
usually the best source for info since the tech docs and manuals are 
lagging behind what the developers are producing. No mention of SQL. 
Hopefully I'm wrong but...

My 6beta2 accounting file still looks 'old school':

> $ more accounting 
> # Version: 6.0Beta2
> # 
> # DO NOT MODIFY THIS FILE MANUALLY!
> # 
> dagqueue:chrisdag.local:dag:dag:hostname:3:sge:0:1083761876:1083761877:1083761877:0:0:0:0:0:0.000000
> :0:0:0:0:0:0:0:1.000000:1:3:138:0:11:0:none:defaultdepartment:none:1:0:0.000000:0.000000:0.000000:-I
>  y:0.000000:none:0.000000
> dagqueue:chrisdag.local:dag:dag:hostname:4:sge:0:1083761892:1083761893:1083761894:0:0:1:0:0:0.000000
> :0:0:0:0:0:0:0:0.000000:0:3:138:0:8:0:none:defaultdepartment:none:1:0:0.000000:0.000000:0.000000:-q 
> dagqueue -I y:0.000000:none:0.000000
> dagqueue:chrisdag.local:dag:dag:hostname:5:sge:0:1083761905:1083761905:1083761905:0:0:0:0:0:0.000000
> :0:0:0:0:0:0:0:0.000000:0:3:138:0:8:0:none:defaultdepartment:none:1:0:0.000000:0.000000:0.000000:-q 
> dagqueue@chrisdag.local -I y:0.000000:none:0.000000
<snip>

That said though, many people have already managed to figure out how to 
parse this file into a mysql database schema. (simple perl script 
basically).

I've always thought that there was a potential money making niche for 
someone who could write a good analytical and reporting package for the 
open source grid engine.

-Chris