Problems after moving the database location

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

Problems after moving the database location

Ewald Grosse-Wilde
Hello,

feeling a bit like an idiot I have decided to ask here for help. We use Apollo for a single genome, running the software on an Azure VM with Ubunbtu. With some help from Nathan I got Apollo 2.5.0 to run, and we had populated it with the initial data. Now some restructuring by our IT meant I had to move the postgresql database to another mount point and followed a guide to do so. This went well, and I could confirm users and database structures were present and accessible via the postgresql client. However, Apollo does not work anymore. It does start and allows users to login. But instead of the view on the left, they get an error that jbrowse couldn't run since it is missing refSeq.json. I have attached a screenshot. From what I can see this directory does not exist, and I assume that is actually normal, and that somehow the database is supposed to 'fill in' for the missing data directory.
I have attached the apollo-config, passwords, and so on removed. Note that the JBrowse part was also changed at one point, with comment tags removed except for the part on myvariantviewer, which I know doesn't work. Leaving the JBrowse part intact did not change anything. I also tried to deploy Apollo 2.6.0 with the same configuration, but unsurprisingly this also didn't change anything. Permissions for apollo, database, and webapp directories are fine, and the mount is mounted with the exact same options as the one we used before (I just replaced the UUID in fstab).
But that also means I have reached the limit of my (sparse) knowledge on how to fix such problems. I am actually 80% certain that I have seen this error before, Iand even should know the fix, which is making me crazy! Can anyone here help me?  

Best,
Ewald
 

--
To unsubscribe from this group and stop receiving emails from it, send an email to [hidden email].
Reply | Threaded
Open this post in threaded view
|

Re: Problems after moving the database location

nathandunn

Ewald, 

I don’t think any of your attachments you referred to came through. 

Offhand, I would try to see if you have just changed the JBrowse organism directory to one you know exists. If it doesn’t let you do that through the organism panel on the organism interface I would update it in the database:

update organism set directory=’new_directory' where common_name = ‘my_organism’;

The Tomcat log might be helpful here, as well. 

Nathan


On Jun 11, 2020, at 12:56 AM, Ewald Grosse-Wilde <[hidden email]> wrote:

Hello,

feeling a bit like an idiot I have decided to ask here for help. We use Apollo for a single genome, running the software on an Azure VM with Ubunbtu. With some help from Nathan I got Apollo 2.5.0 to run, and we had populated it with the initial data. Now some restructuring by our IT meant I had to move the postgresql database to another mount point and followed a guide to do so. This went well, and I could confirm users and database structures were present and accessible via the postgresql client. However, Apollo does not work anymore. It does start and allows users to login. But instead of the view on the left, they get an error that jbrowse couldn't run since it is missing refSeq.json. I have attached a screenshot. From what I can see this directory does not exist, and I assume that is actually normal, and that somehow the database is supposed to 'fill in' for the missing data directory.
I have attached the apollo-config, passwords, and so on removed. Note that the JBrowse part was also changed at one point, with comment tags removed except for the part on myvariantviewer, which I know doesn't work. Leaving the JBrowse part intact did not change anything. I also tried to deploy Apollo 2.6.0 with the same configuration, but unsurprisingly this also didn't change anything. Permissions for apollo, database, and webapp directories are fine, and the mount is mounted with the exact same options as the one we used before (I just replaced the UUID in fstab).
But that also means I have reached the limit of my (sparse) knowledge on how to fix such problems. I am actually 80% certain that I have seen this error before, Iand even should know the fix, which is making me crazy! Can anyone here help me?  

Best,
Ewald
 

--
To unsubscribe from this group and stop receiving emails from it, send an email to [hidden email].
Reply | Threaded
Open this post in threaded view
|

Re: Problems after moving the database location

Ewald Grosse-Wilde
Dear Nathan,

thanks! That actually was the problem - after updating the path manually in the database, everything is back to normal now! Learned something new as well, so thank you also for that!

Best,
Ewald

Am Do., 11. Juni 2020 um 15:54 Uhr schrieb Nathan Dunn <[hidden email]>:

Ewald, 

I don’t think any of your attachments you referred to came through. 

Offhand, I would try to see if you have just changed the JBrowse organism directory to one you know exists. If it doesn’t let you do that through the organism panel on the organism interface I would update it in the database:

update organism set directory=’new_directory' where common_name = ‘my_organism’;

The Tomcat log might be helpful here, as well. 

Nathan


On Jun 11, 2020, at 12:56 AM, Ewald Grosse-Wilde <[hidden email]> wrote:

Hello,

feeling a bit like an idiot I have decided to ask here for help. We use Apollo for a single genome, running the software on an Azure VM with Ubunbtu. With some help from Nathan I got Apollo 2.5.0 to run, and we had populated it with the initial data. Now some restructuring by our IT meant I had to move the postgresql database to another mount point and followed a guide to do so. This went well, and I could confirm users and database structures were present and accessible via the postgresql client. However, Apollo does not work anymore. It does start and allows users to login. But instead of the view on the left, they get an error that jbrowse couldn't run since it is missing refSeq.json. I have attached a screenshot. From what I can see this directory does not exist, and I assume that is actually normal, and that somehow the database is supposed to 'fill in' for the missing data directory.
I have attached the apollo-config, passwords, and so on removed. Note that the JBrowse part was also changed at one point, with comment tags removed except for the part on myvariantviewer, which I know doesn't work. Leaving the JBrowse part intact did not change anything. I also tried to deploy Apollo 2.6.0 with the same configuration, but unsurprisingly this also didn't change anything. Permissions for apollo, database, and webapp directories are fine, and the mount is mounted with the exact same options as the one we used before (I just replaced the UUID in fstab).
But that also means I have reached the limit of my (sparse) knowledge on how to fix such problems. I am actually 80% certain that I have seen this error before, Iand even should know the fix, which is making me crazy! Can anyone here help me?  

Best,
Ewald
 

--
To unsubscribe from this group and stop receiving emails from it, send an email to [hidden email].