v480 Genome Browser Available

[Jonathan Casper]

Good Afternoon Genome Browser Mirror Site Operators:

Version v480 of the UCSC Genome Browser software has now been released.

Please note: We are continuing to build our command-line utilities in a more general environment for ease of use, but those binaries rely on a more recent version of GLIBC.  Due to several requests, we have created a one-time set of frozen binaries using an older GLIBC, which may be suitable for users with older systems.  These binaries can be found on our download server at https://hgdownload.soe.ucsc.edu/admin/exe/linux.x86_64.v479/.

The main changes in this release are:

• Users must now be logged in before using track collections.
• Enabled caching for hubApi.
• Implicitly convert ENST accessions to ENSP accessions during hgvs search if the hgvs term is a p. term.
• Interact tracks gain an option to suppress display of the details boxes. When displayed, those boxes are also larger now - easier to find and click on.
• Revision of the recently-added heatmap display style, with better support for heatmap legends and coloring.
• A thousand groups are now allowed in MAF displays, not just 20.
• hvGfx library fix - hvGfxCurve now correctly relocates points of the curve in reverse complement mode, so hic tracks in arc mode won't look like koosh balls anymore.
• More hubSpace prep for release: add search box for genome selection, row clicks either open the directory or select the row if it was a regular file, wording changes, install hook scripts into cgi-bin/loader, decode or encode file names more appropriately, especially if coming from hubtools, make spaces characters no longer allowed so apache can find the track files easier.
• Fixed memory bug in peakClusters.
• Fixed slNameSort in list.c for hubApi, refreshed expected output lists.
• Fixed bad slNameSort override in findCutters utility.
• AutoSql files that define arrays with a variable number of arguments can now use variables whose name starts with an underscore (obscure and unlikely to be used widely, but it works now).
• bedPartition - add options -minPartitionItems, -partMergeDist, and statistics output. The goal is to produce fewer, larger partitions, so that job startup overhead isn't greater than the runtime.
• Improvements to wuhCor1 big tree build process.

And data-related changes:

• Continuous GenArk updates and liftOver file generation.

For a comprehensive list of changes for this version, please visit/bookmark (you may need to force refresh in your web browser to see an updated page): https://genecats.gi.ucsc.edu/builds/versions.html

We typically release a new version every three weeks. A summary of all past releases and changes can be found at the above link.

The new source code tree is available at:

https://hgdownload.soe.ucsc.edu/admin/jksrc.zip

or labeled with version number at:

https://hgdownload.soe.ucsc.edu/admin/jksrc.v480.zip

or in our Github repository release that is tagged with the version number (https://github.com/ucscGenomeBrowser/kent/releases).

If you use the GBiB virtual machine with auto-updates enabled, it will automatically update itself on Sunday. If you have installed your UCSC Genome Browser server with the GBiC installation script browserSetup.sh, then you can upgrade it with the command "sudo bash browserSetup.sh update".

If you have installed your Genome Browser manually:

• You can use the following rsync command to copy the CGI binaries into the cgi-bin directory of your Apache server: rsync -avP rsync://hgdownload.soe.ucsc.edu/cgi-bin/ ${WEBROOT}/cgi-bin/
• We provide a shell script to update the htdocs directory, htdocs and cgi-bin must be updated together: https://github.com/ucscGenomeBrowser/kent/blob/master/src/product/scripts/updateHtml.sh
• The hgcentral database contains our curated list of public track hubs and pointers to BLAT servers per database and should be updated regularly. We have released a new MySQL dump of this database: https://hgdownload.soe.ucsc.edu/admin/hgcentral.sql

A license is required for commercial download and/or installation of the Genome Browser binaries and source code. No license is needed for academic, nonprofit, and personal use. More information on our licensing page:

http://genome.ucsc.edu/license/

If you have any questions or concerns, please feel free to write back to this genome-mirror mail list.

Jonathan Casper

UCSC Genomics Institute

[Platts, Adrian]

Hey All,

Due to the goings on with CentOS we finally moved to Ubuntu 24.04.2 LTS for our genome browser mirror site.

All seemed to go ok and we were running fine with just a local Zebrafish database until I noticed this morning 

17 pending security updates.

I installed all the updates but the results seem to have been rather undesirable.  First Apache2 was removed

completely - when I couldn’t connect I looked for a stopped service, but there was no apache service at all, indeed 

even the executable had been removed.   Second the mysql directory that was on an external drive and soft 

linked from /var/lib had been completely wiped, also the soft link had gone and had been replaced by an 

empty directory (not a soft linked directory) in /var/lib.

These were both annoying but I installed apache again using apt and purged mysql and mariadb having stopped

the services - thereby allowing the browser setup script to reinstall them.  

But now I hit the really troubling issue.  After the browser setup script has run (just with the install parameter) I

have all the expected tables in mysql … hgcentral etc etc.  But as soon as I move the mysql folder to an 

external drive and soft link from /var/lib these all disappear and I’m left with just  the information-schema 

database.  After that I can’t create databases without getting an error ERROR 1006 (HY000): Can't create database 

(errno: 2 "No such file or directory”).  Even moving the physical mysql directory back to the

root drive doesn’t restore things so I have to uninstall and reinstall mysql/mariadb.

Is this a new security feature or do I somehow need to bless the mysql directory after moving it? Or am I doing something 

stupid like not stopping mysql and mariadb services before moving the mysql directory?  Its potentially troubling as 

our root directory has only 25GB free which isn’t space enough to work offline with any genome.

Adrian

[Hiram Clawson]

Good Morning Adrian:

It is just fine to move MySQL tables around to different filesystems.
When you move the files and leave a symlink to the other location,
there is a MySQL configuration setting you need to add to direct
MySQL to follow the symlinks:

Try adding this to your mysql server configuration:

symbolic-links=1

--Hiram

[Platts, Adrian]

Good Morning Hiram,

Thanks for the suggestion.  While it didn’t immediately work for me, it probably set me looking in the right direction.

I stopped mysql/mariadb and moved the /var/lib/mysql to a new partition then set the symbolic-links paramater in the config 

and tried to start the services back up.  But this failed with what looked like a simple permissions error:

Apr 24 13:53:37 ubuntu systemd[1]: Starting mariadb.service - MariaDB 10.11.11 database server...

Apr 24 13:53:37 ubuntu (mariadbd)[17355]: mariadb.service: Referenced but unset environment variable evaluates to an empty string: MYSQLD_OPTS, _WSREP_NEW_CLUSTER

Apr 24 13:53:37 ubuntu mariadbd[17355]: 2025-04-24 13:53:37 0 [Warning] Can't create test file '/var/lib/mysql/ubuntu.lower-test' (Errcode: 13 "Permission denied”).   <<<

Apr 24 13:53:37 ubuntu mariadbd[17355]: [92B blob data]

Apr 24 13:53:37 ubuntu mariadbd[17355]: 2025-04-24 13:53:37 0 [ERROR] Aborting

Apr 24 13:53:37 ubuntu systemd[1]: mariadb.service: Main process exited, code=exited, status=1/FAILURE

Apr 24 13:53:37 ubuntu systemd[1]: mariadb.service: Failed with result 'exit-code'.

Apr 24 13:53:37 ubuntu systemd[1]: Failed to start mariadb.service - MariaDB 10.11.11 database server.

but the soft linked location had every permission and ownship change I could think of as did the containing location.  In fact I got so paranoid that

I temporarily gave the mysql account a bash shell, su’d into it and checked that I could touch /var/lib/mysql/ubuntu.lower-test

It didn’t make any difference whether the symbolic-link parameter was set or not - mariadb was absolutely confident it couldn’t 

write to the new locatiion and when I tried, I couldn’t create a new database with the soft link in place.

What solved it was starting mysql with the new locatiion explicitly set rather than having it try to believe the soft link, so:

mysqld —datadir=[new location] --user mysql

after that mariadb and mysql were completely cool about reading and writing to the new location even without symbolic-links set

(probably because it isn’t using the symbolic link at all even though the installer script is).

I wonder if the culrit here may be apparmor?  I know on CentOS/redhat the browser install script deactivates SELinux, but I think 

apparmor may be harder to halt?

Adrian

[Maximilian Haeussler]

Hi Adrian,

This does not sound like AppArmor. You could also try to put AppArmor into "complain mode" and check if it was that:

https://wiki.debian.org/AppArmor/HowToUse

 

The internet has a lot of information about how to add directories to AppArmor for Mysqld. The dataDir specifically can be changed to allow MySQL to write to other directories:

https://askubuntu.com/questions/294672/removing-apparmor

best

Max

--

---
You received this message because you are subscribed to the Google Groups "UCSC Genome Browser Mirror-Specific Support" group.
To unsubscribe from this group and stop receiving emails from it, send an email to genome-mirro...@soe.ucsc.edu.
To view this discussion visit https://groups.google.com/a/soe.ucsc.edu/d/msgid/genome-mirror/5FE09AF5-5C7A-420B-973D-51411D7912C7%40msu.edu.