Re: How to get files from crash apps
I am using DEA. But I didn’t found where the container was kept. Is it compressed and stored somewhere? I also took a look at /var/vcap/data/dea_next/crashes/id, it only contain the files in “logs” directory: staging_task.log/stdout.log/stderr.log/env.log, not application files. We need the files under “app” directory. Also I tried to access warden as you said. But the warden under /var/vcap/data/warden/depot was removed immediately after application process was killed.
From: cf-dev-bounces(a)lists.cloudfoundry.org [mailto:cf-dev-bounces(a)lists.cloudfoundry.org] On Behalf Of James Bayer
Sent: 2015年6月24日 16:07
To: Discussions about Cloud Foundry projects and the system overall.
Subject: Re: [cf-dev] How to get files from crash apps
if you're using DEA's i believe the container stays around on the DEA for about an hour or so by default before being removed.
assuming you're on a DEA, to investigate, you'd need to know the DEA and the warden handle to investigate, log in to the DEA as an operator with bosh, then go the warden depot directory, find the warden handle and execute bin/wsh from the container handle directory.
in diego, removing containers happens right away, but diego is also going to support ssh for active troubleshooting and could just increase the memory and attach jvm tools for debugging purposes to troubleshoot.
On Tue, Jun 23, 2015 at 1:12 AM, Meng, Xiangyi <xiangyi.meng(a)emc.com<mailto:xiangyi.meng(a)emc.com>> wrote:
We encountered java out of memory issue in our CF env recently. We had enabled gc log for the application by setting JAVA_TOOL_OPTIONS: -Xloggc:gc.log. But when out of memory issue occurred, the application was stopped and removed. We can’t get gc.log. Is there any way to keep the gc.log after application crashes? The CF version we are using is 197. Any advice would be appreciated.
cf-dev mailing list