You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
/workspace/node-test-commit-aix/nodes/aix61-ppc64/out/Release/.deps//home/iojs/build/workspace/node-test-commit-aix/nodes/aix61-ppc64/out/Release/obj.target/node_base/src/backtrace_posix.o.d.raw -c -o /home/iojs/build/workspace/node-test-commit-aix/nodes/aix61-ppc64/out/Release/obj.target/node_base/src/backtrace_posix.o ../src/backtrace_posix.cc
/home/iojs/build/workspace/node-test-commit-aix/nodes/aix61-ppc64/out/Release/obj/gen/node_javascript.cc:64657:1: fatal error: error writing to /tmp//ccXObGxs.s: No space left on device
} // namespace node
^
One machine looked ok, but the other had a lot of old npm-* directories . Thinking those might be due to cigtm runs. I deleted them and we'll see if that fixes things up or if something has changed and we need more /tmp space
@gibfahn are the directories in the form of 0629edd6-daca-4544-ac99-2ffc3a3f0052 tied to citgm, they seem to be taking a good portion of the space on test-osuosl-aix61-ppc64_be-1 and there are a number of them from yesterday.
@mhdawson yes, those directories are citgm's temp directories. A build was run without the temp directory settings properly configured, which is why they filled up. I thought I'd cleaned up both machines, but I guess not.
You can usually get rid of everything with: rm -rf /tmp/*-*-*-*-* /tmp/npm-*
@gibfahn happened again. Saw a failure over the weekend and tmp was full on test-osuosl-aix61-ppc64_be-2. Lots of directories with watchify-xxxx but those did not contribute significantly to the space.
Issue still seemed to be the directories associated with citgm. Cleaned up all the directories with the command above. I guess we'll see now if it continues to pile up.
Activity
mhdawson commentedon Apr 3, 2017
One machine looked ok, but the other had a lot of old npm-* directories . Thinking those might be due to cigtm runs. I deleted them and we'll see if that fixes things up or if something has changed and we need more /tmp space
mhdawson commentedon Apr 3, 2017
Noticed that deleting npm-* only freed up 2%. Looks like this one is the bigger culprit:
mhdawson commentedon Apr 3, 2017
deleting that freed up about 50%
mhdawson commentedon Apr 3, 2017
@gibfahn are the directories in the form of 0629edd6-daca-4544-ac99-2ffc3a3f0052 tied to citgm, they seem to be taking a good portion of the space on test-osuosl-aix61-ppc64_be-1 and there are a number of them from yesterday.
mhdawson commentedon Apr 3, 2017
Build to see if I've freed up enough space: https://ci.nodejs.org/job/node-test-commit-aix/4800/
gibfahn commentedon Apr 4, 2017
@mhdawson yes, those directories are citgm's temp directories. A build was run without the temp directory settings properly configured, which is why they filled up. I thought I'd cleaned up both machines, but I guess not.
You can usually get rid of everything with:
rm -rf /tmp/*-*-*-*-* /tmp/npm-*mhdawson commentedon Apr 4, 2017
can you take a look and clean up the rest. When I looked yesterday I only cleaned up enough to get the machines working again.
mhdawson commentedon Apr 10, 2017
@gibfahn happened again. Saw a failure over the weekend and tmp was full on test-osuosl-aix61-ppc64_be-2. Lots of directories with watchify-xxxx but those did not contribute significantly to the space.
Issue still seemed to be the directories associated with citgm. Cleaned up all the directories with the command above. I guess we'll see now if it continues to pile up.
richardlau commentedon Apr 11, 2017
#682
Trott commentedon Apr 11, 2017
Just ran
rm -rf /tmp/*-*-*-*-* /tmp/npm-*again now and that seems to have cleared things up. Obviously, needing a more permanent solution...gibfahn commentedon Apr 11, 2017
See nodejs/citgm#396 (comment) , this shouldn't be happening if jobs are configured correctly.
sam-github commentedon Aug 13, 2019
closing, stale issue.
current state is: