From ziscloud at gmail.com Wed Aug 19 10:04:27 2020 From: ziscloud at gmail.com (=?UTF-8?B?546L6aG65LqR?=) Date: Wed, 19 Aug 2020 03:04:27 -0700 Subject: softreference, 0 refs report high cost time Message-ID: Hi Forks, I have an application that reported high (more than 1 second) times to process *zero* SoftReferences, for example: 2020-08-11T18:58:19.222+0800: 611331.604: [SoftReference, 0 refs, 1.4741673 secs]2020-08-11T18:58:20.696+0800: 611333.079: [WeakReference, 0 refs, 0.0004746 secs]2020-08-11T18:58:20.696+0800: 611333.079: [FinalReference, 4559 refs, 0.0019431 secs]2020-08-11T18:58:20.698+0800: 611333.081: [PhantomReference, 2 refs, 0 refs, 0.0016303 secs]2020-08-11T18:58:20.700+0800: 611333.083: [JNI Weak Reference, 0.0000284 secs], 1.4867549 secs] 2020-08-14T02:58:22.222+0800: 812934.605: [SoftReference, 0 refs, 1.0247147 secs]2020-08-14T02:58:23.247+0800: 812935.629: [WeakReference, 1 refs, 0.0007859 secs]2020-08-14T02:58:23.247+0800: 812935.630: [FinalReference, 4382 refs, 0.0024934 secs]2020-08-14T02:58:23.250+0800: 812935.633: [PhantomReference, 2 refs, 0 refs, 0.0008613 secs]2020-08-14T02:58:23.251+0800: 812935.634: [JNI Weak Reference, 0.0000239 secs], 1.0380333 secs] We have been hit by this anomaly a few times now, and detailed information can be find in the attached logs (that also show the command line flags). OS information: Linux 3.10.0-862.el7.x86_64 x86_64 NAME="CentOS Linux" VERSION="7 (Core)" Docker version: 18.09.9 Thanks for the help. Tony -------------- next part -------------- An HTML attachment was scrubbed... URL: From stefan.johansson at oracle.com Fri Aug 21 11:49:23 2020 From: stefan.johansson at oracle.com (stefan.johansson at oracle.com) Date: Fri, 21 Aug 2020 13:49:23 +0200 Subject: softreference, 0 refs report high cost time In-Reply-To: References: Message-ID: <7cae0f60-7b82-e886-c079-d9e1f8674791@oracle.com> Hi, None of your attachments made it through to the lists. Try subscribing to the list if you haven't already done that, it should help. Judging from the log snippets in this initial mail you seem to be running a fairly old JDK. Have you trying with JDK 14 to see if the problems still exists in a recent release? There have been a lot of improvements to reference processing the last couple of years. Thanks, Stefan On 2020-08-19 12:04, ??? wrote: > Hi Forks, > > I have an application that reported high (more than 1 second) times to > process *zero* SoftReferences, for example: > > 2020-08-11T18:58:19.222+0800: 611331.604: [SoftReference, 0 refs, > 1.4741673 secs]2020-08-11T18:58:20.696+0800: 611333.079: [WeakReference, > 0 refs, 0.0004746 secs]2020-08-11T18:58:20.696+0800: 611333.079: > [FinalReference, 4559 refs, 0.0019431 secs]2020-08-11T18:58:20.698+0800: > 611333.081: [PhantomReference, 2 refs, 0 refs, 0.0016303 > secs]2020-08-11T18:58:20.700+0800: 611333.083: [JNI Weak Reference, > 0.0000284 secs], 1.4867549 secs] > > 2020-08-14T02:58:22.222+0800: 812934.605: [SoftReference, 0 refs, > 1.0247147 secs]2020-08-14T02:58:23.247+0800: 812935.629: [WeakReference, > 1 refs, 0.0007859 secs]2020-08-14T02:58:23.247+0800: 812935.630: > [FinalReference, 4382 refs, 0.0024934 secs]2020-08-14T02:58:23.250+0800: > 812935.633: [PhantomReference, 2 refs, 0 refs, 0.0008613 > secs]2020-08-14T02:58:23.251+0800: 812935.634: [JNI Weak Reference, > 0.0000239 secs], 1.0380333 secs] > > We have been hit by this anomaly a few times now, and ?detailed > information can be find in the attached logs (that also show the command > line flags). > > OS information:?Linux 3.10.0-862.el7.x86_64 x86_64 > NAME="CentOS Linux" > VERSION="7 (Core)" > > Docker version: > 18.09.9 > > Thanks for the help. > > Tony > > _______________________________________________ > hotspot-gc-use mailing list > hotspot-gc-use at openjdk.java.net > https://mail.openjdk.java.net/mailman/listinfo/hotspot-gc-use >