Re: [controller-dev] [integration-dev] ODL crashing in CSIT jobs

2017-11-29 Thread Thanh Ha
We can certainly add swap but imo swap is just a slower ram. At least in my experience when a system starts needing to swap things have already gone beyond recovery and swap just delays the inevitable program crash. In any case there's a few ways we can add swap if we really want to: 1) At the job

Re: [controller-dev] odlparent 3.0.2 for CONTROLLER-1799: Archetype self test during Maven build

2017-11-29 Thread Michael Vorburger
On Mon, Nov 27, 2017 at 8:10 PM, Michael Vorburger wrote: > Hello, > > almost 3 months ago, on https://lists.opendaylight. > org/pipermail/controller-dev/2017-September/013889.html, I had started a > thread re. the mysterious problems hit in the controller archetype "self > test", which occured o

Re: [controller-dev] [integration-dev] ODL crashing in CSIT jobs

2017-11-29 Thread Luis Gomez
I do not remember all the issues related with memory but at least swap is required in the robot VM to generate the log.html when output.xml file is big. > On Nov 29, 2017, at 2:56 AM, Thanh Ha wrote: > > We can certainly add swap but imo swap is just a slower ram. At least in my > experience w

Re: [controller-dev] [integration-dev] ODL crashing in CSIT jobs

2017-11-29 Thread Jamo Luhrsen
Well, I think Thanh's point is that if the output.xml is too big, we just add more RAM to fix it, although instead we could add some swap which is "slower RAM" JamO On 11/29/2017 09:13 AM, Luis Gomez wrote: > I do not remember all the issues related with memory but at least swap is > required i

Re: [controller-dev] [integration-dev] ODL crashing in CSIT jobs

2017-11-29 Thread Luis Gomez
OK, if we have RAM lets add it, I thought we had some constraint there. > On Nov 29, 2017, at 9:43 AM, Jamo Luhrsen wrote: > > Well, I think Thanh's point is that if the output.xml is too big, we just add > more RAM to > fix it, although instead we could add some swap which is "slower RAM" > >

Re: [controller-dev] [integration-dev] ODL crashing in CSIT jobs

2017-11-29 Thread Jamo Luhrsen
actually, I get your point, Luis. There is always a little give and take when we want to ask for more RAM. I take it that swap space is not going to be a big deal since storage is not something we are every worried about. right Thanh? btw, we did bump the robot builder vms from 2G to 4G RAM rec

Re: [controller-dev] [integration-dev] ODL crashing in CSIT jobs

2017-11-29 Thread Thanh Ha
Yes, that's right the systems have 4GB of RAM now. If the only things we need swap for is the robot systems then I'm more inclined to suggest we create it as part of the job definition first, and move it to packer if we find it adds to much build time (I'm not sure if creating a swap file is a sign