[
https://issues.apache.org/jira/browse/CASSANDRA-12699?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Ariel Weisberg updated CASSANDRA-12699:
---------------------------------------
Attachment: cassandraMemoryLog.sh
The version of the memory logging script you provided has windows line endings
and didn't run for me. I fixed the line endings and added VmData as well. I
find VmRSS to not be very useful because it includes cached data that can be
evicted or flushed.
I am looking at issues with the OOM killer as well, and I happen to just have
tested this. You are seeing 10 gigabytes used by VmData which in my digging is
basically anonymous memory. I was able to track it down to mapped anonymous
memory in /proc/pid/maps in my case.
That amount of PTE usage doesn't seem insane to me if it has to map 3
terabytes, although sure less is better. If you want to manage more data you
need more memory. Maybe the real problem is that PTE usage can't temporarily
spike because all the memory is committed.
There is a lot of hard to explain off heap memory utilization in C*.
java.nio.Bits.java reports only 1 gigabyte or so off heap memory in my case and
bloom filters only account for another gigabyte, but I see 10 gigabytes or so
in VmData that I can't account for and that amount grows until the OOM killer
kicks in. I don't run out of space for PTEs before that happens.
> Excessive use of "hidden" Linux page table memory
> -------------------------------------------------
>
> Key: CASSANDRA-12699
> URL: https://issues.apache.org/jira/browse/CASSANDRA-12699
> Project: Cassandra
> Issue Type: Bug
> Components: Core
> Environment: Cassandra 2.2.7 on Red Hat 6.7, with Java 1.8.0_73.
> Probably others.
> Reporter: Heiko Sommer
> Attachments: PageTableMemoryExample.png, cassandra-env.sh,
> cassandra.yaml, cassandraMemoryLog.sh, cassandraMemoryLog.sh
>
>
> The cassandra JVM process uses many gigabytes of page table memory during
> certain activities, which can lead to oom-killer action with
> "java.lang.OutOfMemoryError: null" logs.
> Page table memory is not reported by Linux tools such as "top" or "ps" and
> therefore might be responsible also for other spurious Cassandra issues with
> "memory eating" or crashes, e.g. CASSANDRA-8723.
> The problem happens especially (or only?) during large compactions and
> anticompactions.
> Eventually all memory gets released, which means there is no real leak. Still
> I suspect that the memory mappings that fill the page table could be released
> much sooner, to keep the page table size at a small fraction of the total
> Cassandra process memory.
> How to reproduce: Record the memory use on a Cassandra node, including page
> table memory, for example using the attached script cassandraMemoryLog.sh.
> Even when there is no crash, the ramping up and sudden release of page table
> memory is visible.
> A stacked area plot for the memory on one of our crashed nodes is attached
> (PageTableMemoryExample.png). The page table memory used by Cassandra is
> shown in red ("VmPTE").
> (In the plot we also see that the sum of measured memory portions sometimes
> exceeds the total memory. This is probably an issue of how RSS memory is
> measured, perhaps including some buffers/cache memory that also counts toward
> available memory. It does not invalidate the finding that page table memory
> is growing to enormous sizes.)
> Shortly before the crash, /proc/$PID/status reported
> VmPeak: 6989760944 kB
> VmSize: 5742400572 kB
> VmLck: 4735036 kB
> VmHWM: 8589972 kB
> VmRSS: 7022036 kB
> VmData: 10019732 kB
> VmStk: 92 kB
> VmExe: 4 kB
> VmLib: 17584 kB
> VmPTE: 3965856 kB
> VmSwap: 0 kB
> The files cassandra.yaml and cassandra-env.sh used on the node where the data
> was taken are attached.
> Please let me know if I should provide any other data or descriptions to help
> with this ticket.
> Known workarounds: Use more RAM, or limit the amount of Java heap memory. In
> the above crash, MAX_HEAP_SIZE was not set, so that the default heap size for
> 12 GB RAM was used (-Xms2976M, -Xmx2976M).
> We have not tried yet if variations of heap vs. offheap config choices make a
> difference.
> Perhaps there are other workarounds using -XX+UseLargePages or related Linux
> settings to reduce the size of the process page table?
> I believe that we see these crashes more often than other projects because we
> have a test system with not much RAM but with a lot of data (compressed ~3 TB
> per node), while the CPUs are slow so that anti-/compactions overlap a lot.
> Ideally Cassandra (native) code should be changed to release memory in
> smaller chunks, so that page table size cannot cause an otherwise stable
> system to crash.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
