System unavailable: Solaris X86_64 Nodes - Unavailable

[steelhead31]

The 2 solaris X86_64 nodes are both unusable

build-packet_esxi-solaris10u11-x64-1 is out of memory ( log below )

09:35:20 # There is insufficient memory for the Java Runtime Environment to continue.
09:35:20 # Native memory allocation (mmap) failed to map 243269632 bytes for committing reserved memory.
From Grinder

build-packet_esxi-solaris10u11-x64-2 is offline

[sxa]

Related: #2441 which mentions -2 being offline

[sxa]

There are 26836 defunct processes on the -1 machine. 1351 of those under the jenkins ID. I'll reboot it.

[sxa]

Jenkins process listing (excluding the defunct ones): build-packet_esxi-solaris10u11-x64.1.leftovers.txt.gz

There are quite a lot from old jobs which is likely exhausting the resources. See also: #2442 which would potentially have caught many of these
I've triggered a reboot

[sxa]

-1 is back and currently showing a load everage of, erm, 8388608.00.
Which I assume is fake news as it seems pretty responsive and the jenkins agent has been launched successfully on it

[steelhead31]

Once I've got access, I'll have a look at fixing #2442

[sxa]

Solaris x64 machines were all rebuilt as part of #2792 #2666 and I'm not aware of any such issues on the new ones, therefore closing.