Question: Local Galaxy Server on IBM System x3550 ?
gravatar for fire_water
4.5 years ago by
United States
fire_water10 wrote:


Our lab has inherited a server on which we would like to install a local Galaxy production server so that the labs in our department can connect to it and run various jobs. The specifics of the server are as follows:

  • IBM System x3550
  • processors: 2 x Intel Xeon 5140 2.33 GHz
  • memory: 8 GB
  • disks: 2 x 300 GB
  • additional storage is possible via USB

My questions are:

  • is this machine capable of serving as a Galaxy server for a hand full of people?
  • which operating system is appropriate for this machine? Our initial thoughts are Ubuntu Linux server edition.
  • are there pitfalls I should be aware of if we choose to install Galaxy on this machine?

Update 1:

Our lab members will mostly be using bioinformatics tools such as

  • bowtie
  • tophat
  • cufflinks, cuffdif, etc.

to process RNA sequencing data.

Thank you :)

galaxy server • 1.2k views
ADD COMMENTlink modified 4.5 years ago • written 4.5 years ago by fire_water10
gravatar for Martin Čech
4.5 years ago by
Martin Čech ♦♦ 4.9k
United States
Martin Čech ♦♦ 4.9k wrote:

I will point someone more knowledgable than me here, but my 2c for now:

  • The capability of the machine completely depends on the type and number of jobs you want to run, Galaxy itself is not resource-hungry. This said you are mostly limited by the space (600GB - depends on what you analyze again)
  • Ubuntu is fine, from my experience CentOS is also good.


As of pitfalls the good start for learning how to set up production Galaxy is here:

Regarding the job types you want to run the best place to search the resource consumption of these would probably be the tool websites itself (e.g. bowtie: is an ultrafast, memory-efficient short read aligner. It aligns short DNA sequences (reads) to the human genome at a rate of over 25 million 35-bp reads per hour. Bowtie indexes the genome with a Burrows-Wheeler index to keep its memory footprint small: typically about 2.2 GB for the human genome (2.9 GB for paired-end).) source:

Another way how to test the memory footprint is to run typical job separately via commandline and watch the memory. As stated above the fact it is run through Galaxy will not add to that.

ADD COMMENTlink modified 4.5 years ago • written 4.5 years ago by Martin Čech ♦♦ 4.9k

@Martin: thank you for your help. I updated my question to include the types of bioinformatics tools we mostly use.

ADD REPLYlink written 4.5 years ago by fire_water10
Please log in to add an answer.


Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 16.09
Traffic: 172 users visited in the last hour