search

si-medbif / exomeseq

Analysing sequence data from whole exome sequencing or targeted gene panel sequencing.
2 stars 1 forks source link

Memory problem in DigitialOcean #8

Closed dummai closed 6 years ago

dummai commented 6 years ago

The following error was encountered when the pipeline was run on DigitalOcean droplet (4vCPU, 8GB RAM, 160 GB HD):

There is insufficient memory for the Java Runtime Environment to continue. Native memory allocation (mmap) failed to map 120061952 bytes for committing reserved memory. OpenJDK 64-Bit Server VM warning: INFO: os::commit_memory(0x00000006ea300000, 120061952, 0) failed; error='Cannot allocate memory' (errno=12) An error report file with more information is saved as: /usr/hs_err_pid1.log

haraldgrove commented 6 years ago

It's possible to see the total memory using something like "free", maybe we should add a check for this during setup?

dummai commented 6 years ago

I have reduced "java_mem" from 8GB to 7GB. This is a temporary patch to make sure the pipeline will run in a DigitalOcean droplet. Please let me know if you have a better way to fix this.

It's possible to see the total memory using something like "free", maybe we should add a check for this during setup?

hypotheses commented 6 years ago

I think the total memory 8Gb doesnt imply the total useable memory. So, Ton’s solution is probably the best one now.

On Wed, 12 Sep 2018 at 09:57 DumMai notifications@github.com wrote:

I have reduced "java_mem" from 8GB to 7GB. This is a temporary patch to make sure the pipeline will run in a DigitalOcean droplet. Please let me know if you have a better way to fix this.

It's possible to see the total memory using something like "free", maybe we should add a check for this during setup?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/si-medbif/exomeseq/issues/8#issuecomment-420495422, or mute the thread https://github.com/notifications/unsubscribe-auth/AAdOd8yTmeAXQBVD5rp0saa9gxXH_KgGks5uaHgQgaJpZM4WixkI .

-- Best, ~Bhoom Sent from Gmail Mobile

haraldgrove commented 6 years ago

I'll close this now, it probably makes some sense to say that the pipeline requires 8 GB to run properly, which equates to 7 GB for the pipeline and 1 GB for system activity. Increasing the value when the user has more memory available is probably not something we should be doing automatically anyway.