DB user access, trackDB

4 views
Skip to first unread message

Lachlan Musicman

unread,
Apr 15, 2015, 11:37:55 AM4/15/15
to genome...@soe.ucsc.edu, Pip Griffin
Hi,

I am wondering which databases are the explicitly needed data bases
for a new installation that isn't planning on using any of the UCSC
data.

For initial set up I used a combination of some blogs and Minimal
Browser Installation
http://genomewiki.ucsc.edu/index.php/Minimal_Browser_Installation

I haven't seen a minimum required list of databases, although I've
gleaned that hgcentral is the only one that is absolutely required -
until now. (see below, I'm building to it)

I have been following this description for the genome DBs:

http://genomewiki.ucsc.edu/index.php/Building_a_new_genome_database

Interestingly, at no point in the "Building a new genome database"
process (http://genomewiki.ucsc.edu/index.php/Building_a_new_genome_database)
is "giving readonly access to that database" mentioned. I will be
adding it to my automation script, which I hope to share as soon as
it's satisfactorily complete.

I checked the ex.MySQLUserPerms.sh and note that:

#
# Read only access to genome databases for the browser CGI binaries
#
for DB in cb1 hgFixed hg17 proteins040315
do
${MYSQL} -e "GRANT SELECT, CREATE TEMPORARY TABLES on \
${DB}.* TO readonly@localhost IDENTIFIED BY 'access';" mysql
done


Note that I have none of the databases mentioned in the above stanza.
Also, there is no step within the "Building a new database" that
indicates that db users should be provided access to the new dbs in
order for them to be read. I think this needs addressing in the docs,
unless I'm missing something?

Now that I am at Step 13 of the "Building a new database" wiki, it
would seem that I also need a trackDB database.

The documentation is very clear on how to add multiple, different,
trackDB databases to the hg.conf, but it isn't very clear on why you
would be interested in doing that at all, nor at what stage in the
installation these databases are created. Also, which db users need
access to that/those databases?

So, the instructions for the creation of that database seem to be
missing, although there seems to be a couple of instructions on how to
add to that database, they are disparate and in some cases, different.

A single, clear, how to would be really useful.





Cheers
L.



------

I'm treading carefully
but it's the time of night
the snowy light
the subway roar
and the whispered fights
exciting sights
but it's not enough
I thought it was
I wish it was
I thought it was
-----
You name it - The Cannanes with Explosion Robinson.
https://youtu.be/MGrillKAzsI

Lachlan Musicman

unread,
Apr 15, 2015, 11:39:05 AM4/15/15
to genome...@soe.ucsc.edu
Woah.

TrackDB is a table on my genome db rather than a db in its own right?

IE, I will create a trackDb table in droBir1 (and every other genome we add)?


cheers
L.
------

I'm treading carefully
but it's the time of night
the snowy light
the subway roar
and the whispered fights
exciting sights
but it's not enough
I thought it was
I wish it was
I thought it was
-----
You name it - The Cannanes with Explosion Robinson.
https://youtu.be/MGrillKAzsI


Luvina Guruvadoo

unread,
Apr 16, 2015, 2:24:08 PM4/16/15
to Lachlan Musicman, genome...@soe.ucsc.edu
Hello Lachlan,

Thanks for your question. Yes, trackDb is a table. Each assembly (database) will have its own trackDb table. Adding multiple trackDb tables to the hg.conf listing is useful when you have multiple sources of trackDb data. For example, you might want to download UCSC's trackDb while also having your own additional tracks. Having multiple trackDb tables makes that easier. See http://genome-source.cse.ucsc.edu/gitweb/?p=kent.git;a=blob_plain;f=src/product/README.trackDb;hb=HEAD for more information, including some directions for defining your own trackDb for your own assembly.

I hope this helps. If you have any further questions, please reply to gen...@soe.ucsc.edu. All messages sent to that address are archived on a publicly-accessible forum. If your question includes sensitive data, you may send it instead to genom...@soe.ucsc.edu.

- - -
Luvina Guruvadoo
UCSC Genome Bioinformatics Group


--


Reply all
Reply to author
Forward
0 new messages