Re: [MirageOS-devel] Memory exhaustion in Mirage

[Sbstn Mlr <muellerseg@xxxxxxxxx>]

Hi,

I have encountered a similar problem. Here a small example (echoes ethernet frames) running on xen:

open Lwt

module Main (C: V1_LWT.CONSOLE) (N: V1_LWT.NETWORK) = struct

    let start console network =

        let acceptPacket buffer =

            N.write network buffer

        in

        N.listen network acceptPacket

end

This example works fine for low traffic rates with several thousand packets per second.

If I increase the packet rate to 50000 packets per second the program starts to drop packets. After a few seconds this message appears several times:Â

Cannot handle page request order 1!

memalign(4096, 4096) failed.

And the program stops working.Â

Cheers,

S.

2014-12-17 13:41 GMT+01:00 Anil Madhavapeddy <anil@xxxxxxxxxx>:

That's correct. It's pinging from dom0 to the unikernel domU.

-anil

> On 17 Dec 2014, at 04:34, Richard Mortier <Richard.Mortier@xxxxxxxxxxxxxxxx> wrote:
>
> to check-- that's flood ping from a standard ping on linux hitting a unikernel, rather than flood pinging from within a unikernel to something else, correct?
>
> (don't know if it's significant but might help masoud in tracking down his issue.)
>
> On 17 Dec 2014, at 11:32, Anil Madhavapeddy <anil@xxxxxxxxxx> wrote:
>
>> And just to confirm that we have no packet loss, I left parallel ping floods running for a day. The variance is due to the GC compactions:
>>
>> 478535022 packets transmitted, 478535022 received, 0% packet loss, time 42469559ms
>> rtt min/avg/max/mdev = 0.025/0.051/5.549/0.073 ms, ipg/ewma 0.088/0.065 ms
>>
>> -anil
>>
>>> On 16 Dec 2014, at 15:59, Anil Madhavapeddy <anil@xxxxxxxxxx> wrote:
>>>
>>> If you're running short of memory on 2GB, that indicates that there's a large amount of buffering going on. Are you seeing packet loss as a result?
>>>
>>> Please do make a note on the bug report that you're seeing this, and what the trigger conditions are. If you have a simple mirage-skeleton example of an Ethernet port flood that causes this, then that'll help fix it.
>>>
>>> -anil
>>>
>>>> On 16 Dec 2014, at 15:57, Masoud Koleini <masoud.koleini@xxxxxxxxxxxxxxxx> wrote:
>>>>
>>>> Yes, I have faced this problem several times when flooding the input port (similar error on the console). The allocated memory for the switch is 2GB, which is much higher than 64MB, but I guess bit rate should be higher too.
>>>>
>>>> On 16/12/14 23:45, Anil Madhavapeddy wrote:
>>>>> On 16 Dec 2014, at 15:23, Masoud Koleini <masoud.koleini@xxxxxxxxxxxxxxxx> wrote:
>>>>>> I am wondering if the following issue is already addressed:
>>>>>>
>>>>>> https://github.com/mirage/mirage-tcpip/issues/33
>>>>>>
>>>>>> https://lists.cam.ac.uk/pipermail/cl-mirage/2013-August/msg00104.html
>>>>>>
>>>>> It probably still happens at 64MB of RAM -- it requires some code to adjust
>>>>> the GC parameters in the OCaml runtime to trigger a collection more often.
>>>>> It should be harmless however, since (as the bug report observes), a failure
>>>>> to allocate an Io_page results in a GC compaction that frees up memory so
>>>>> that the allocation eventually succeeds.
>>>>>
>>>>> Any particular reason for asking -- is the bug affecting your switch somehow?
>>>>>
>>>>> -anil
>>>>
>>>>
>>>>
>>>>
>>>>
>>>> This message and any attachment are intended solely for the addressee and may contain confidential information. If you have received this message in error, please send it back to me, and immediately delete it. ÂPlease do not use, copy or disclose the information contained in this message or in any attachment. Any views or opinions expressed by the author of this email do not necessarily reflect the views of the University of Nottingham.
>>>>
>>>> This message has been checked for viruses but the contents of an attachment may still contain software viruses which could damage your computer system, you are advised to perform your own checks. Email communications with the University of Nottingham may be monitored as permitted by UK legislation.
>>>>
>>>
>>>
>>> _______________________________________________
>>> MirageOS-devel mailing list
>>> MirageOS-devel@xxxxxxxxxxxxxxxxxxxx
>>> http://lists.xenproject.org/cgi-bin/mailman/listinfo/mirageos-devel
>>>
>>
>>
>> _______________________________________________
>> MirageOS-devel mailing list
>> MirageOS-devel@xxxxxxxxxxxxxxxxxxxx
>> http://lists.xenproject.org/cgi-bin/mailman/listinfo/mirageos-devel
>
>
> --
> Cheers,
>
> R.
>
> [ This address fails on Dec31. Use richard.mortier@xxxxxxxxxxxx subsequently. ]
>
>
>
>
>
>
>
> This message and any attachment are intended solely for the addressee and may contain confidential information. If you have received this message in error, please send it back to me, and immediately delete it. ÂPlease do not use, copy or disclose the information contained in this message or in any attachment. Any views or opinions expressed by the author of this email do not necessarily reflect the views of the University of Nottingham.
>
> This message has been checked for viruses but the contents of an attachment may still contain software viruses which could damage your computer system, you are advised to perform your own checks. Email communications with the University of Nottingham may be monitored as permitted by UK legislation.
>


_______________________________________________
MirageOS-devel mailing list
MirageOS-devel@xxxxxxxxxxxxxxxxxxxx
http://lists.xenproject.org/cgi-bin/mailman/listinfo/mirageos-devel

_______________________________________________
MirageOS-devel mailing list
MirageOS-devel@xxxxxxxxxxxxxxxxxxxx
http://lists.xenproject.org/cgi-bin/mailman/listinfo/mirageos-devel