Re: jmx-dev RFR: 8226575: OperatingSystemMXBean should be made container aware

2019-12-10 Thread David Holmes
On 10/12/2019 5:31 am, Daniil Titov wrote: Hi David, So we never try to access the uninitialized counters.cpus array which is good but we still return garbage for counters.jvmTicks and counters.cpuTicks - surely that should have been noticeable? It only affected the first time the

Re: jmx-dev RFR: 8226575: OperatingSystemMXBean should be made container aware

2019-12-10 Thread Daniil Titov
Hi David, > Please file a follow up RFE to look into this. I created an issue to follow this up [1] [1] https://bugs.openjdk.java.net/browse/JDK-8235681 Thank you, Daniil On 12/10/19, 2:11 AM, "David Holmes" wrote: On 10/12/2019 5:31 am, Daniil Titov wrote: > Hi David, > >

Re: jmx-dev RFR: 8226575: OperatingSystemMXBean should be made container aware

2019-12-10 Thread Daniil Titov
Hi Serguei, > Do we need to check if the (limit - memLimit) value is negative? > The same question is for getFreeSwapSpaceSize(): > memSwapLimit - memLimit - (memSwapUsage - memUsage) > > and getFreeMemorySize(): >101 return limit - usage; I don't think we nee