Problem Launching Cloud Instance

Question: Problem Launching Cloud Instance

7.1 years ago by

Clare Sloggett • 120 wrote:

Hi all, I wasn't sure whether to send this to galaxy-dev or galaxy-user. I'm having a problem launching a Galaxy instance on Amazon. I have done this in the past with no problems. I suspect I'm doing something trivial wrong but can't work out what! The problem is that after starting the instance, when I go to http://public-dns-url/cloud and the "Initial Cluster Configuration" window pops up, the initial storage size text box is disabled. That is, the text box after the first radio box ("Galaxy Cluster") is greyed out. So, I can't enter a size for the storage and if I just proceed without doing so, it (I think) causes CloudMan to fail. I thought I was probably just having browser issues but I get the same problem repeatedly in Chrome and Firefox. When I inspect the element, I see <input disabled="true" ...="" name="g_pss" ...=""> so I think it really is disabled. This seems like a weird problem so I suspect it's me, but I can't work it out. Any help would be appreciated! Thanks, Clare -- E: sloc@unimelb.edu.au P: 03 903 53357 M: 0414 854 759

galaxy • 828 views

ADD COMMENT • link •

modified 7.1 years ago • written 7.1 years ago by Clare Sloggett • 120

7.1 years ago by

Clare Sloggett • 120

Clare Sloggett • 120 wrote:

Update - I think I can see the problem, and I think it's a bug in the image on Amazon. I might have misunderstood, but thought I would point this out as a possible cause, assuming others experience the same issue I do. Looking at the repository at https://bitbucket.org/galaxy/cloudman/src/65f0ff532a29/templates/index .mako I can see GB... It looks like the input is being set to disabled in all cases instead of just when (as the comments suggest) the underlying machine image doesn't support it? It does look like this is a recent change - I think the culprit is commit https://bitbucket.org/galaxy/cloudman/changeset/65f0ff532a29 #chg-templates/index.mako The thing I don't understand though is that I thought the image I was launching was older than this change. I launched 861460482541/galaxy-cloudman-2011-03-22 , and another older image seems to have the same problem . Do the instances draw on a configuration script which is newer than the image? Or maybe the cause is not what I've suggested! Any suggestions for quick workarounds to get an instance running successfully would also very much appreciated as I was planning to use a CloudMan instance to demo both Galaxy and CloudMan tomorrow to some people interested in using them (in about 11 hours from now in fact). Could I maybe log in to the CloudMan back end and configure it from there? Thanks, Clare -- E: sloc@unimelb.edu.au P: 03 903 53357 M: 0414 854 759

ADD COMMENT • link written 7.1 years ago by Clare Sloggett • 120

Clare; Enis will be able to better answer the questions about fixing this long term, but here is a quick workaround to get it running immediately: ssh to the image and edit '/mnt/cm/templates/index.mako' to remove "disabled='true'" from this line: https://bitbucket.org/galaxy/cloudman/src/b017694e2cfa/templates/index .mako#cl-384 Then if you refresh the http://your-dns/cloud page you'll be able to enter into this box. From there everything was working fine for me. Sorry about the problem and thanks for bringing this up, Brad

ADD REPLY • link written 7.1 years ago by Brad Chapman • 240

Hi Clare, Sorry about the trouble. I just disabled that if statement in the template so the option should no longer be disabled for any new cluster. My goal was to simplify usage but I'll obviously have to revisit the implementation... Sorry for the trouble and good luck with the demo tomorrow. Let me know if you run in to any other issues as an update to CloudMan was released just a couple of days ago. Enis

ADD REPLY • link written 7.1 years ago by Enis Afgan • 680

Wow, thanks for the fast response Brad & Enis!

ADD REPLY • link written 7.1 years ago by Clare Sloggett • 120

Similar posts • Search »