Am 19.11.2012 16:22, schrieb Paolo Bonzini:
Il 19/11/2012 16:04, Stefan Priebe - Profihost AG ha scritto:
[ 49.183366] sd 2:0:0:1: [sdb]
[ 49.183366] Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
[ 49.183366] sd 2:0:0:1: [sdb]
[ 49.183366] Sense Key : Aborted Command [current]
[ 49.1
Il 19/11/2012 16:04, Stefan Priebe - Profihost AG ha scritto:
> [ 49.183366] sd 2:0:0:1: [sdb]
> [ 49.183366] Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
> [ 49.183366] sd 2:0:0:1: [sdb]
> [ 49.183366] Sense Key : Aborted Command [current]
> [ 49.183366] sd 2:0:0:1: [sdb]
> [ 49.183
Hi Paolo,
new patch attached. Desciption is still wrong.
> I think this is all unneeded. Just store rcb->ret into
> rcb->acb->status, and your version of qemu_rbd_aio_cancel should just
> work.
>
> Also, I think the acb->cancelled field is not necessary anymore after
> these changes.
1.) It
Il 19/11/2012 15:48, Stefan Priebe - Profihost AG ha scritto:
> So you mean in qemu_rbd_complete_aio i should remove the check for
> cancelled and then just overwrite acb->status to that it changes from
> -EINPROGRESS to something else?
>
> And qemu_rbd_aio_cancel should just wait for status != -E
Am 19.11.2012 15:41, schrieb Paolo Bonzini:
Il 19/11/2012 15:28, Stefan Priebe - Profihost AG ha scritto:
typedef struct RADOSCB {
@@ -376,6 +377,10 @@ static void qemu_rbd_complete_aio(RADOSCB *rcb)
RBDAIOCB *acb = rcb->acb;
int64_t r;
+if (acb->bh) {
+return;
+}
Il 19/11/2012 15:28, Stefan Priebe - Profihost AG ha scritto:
> typedef struct RADOSCB {
> @@ -376,6 +377,10 @@ static void qemu_rbd_complete_aio(RADOSCB *rcb)
> RBDAIOCB *acb = rcb->acb;
> int64_t r;
>
> +if (acb->bh) {
> +return;
> +}
> +
> r = rcb->ret;
>
>
Il 19/11/2012 15:16, Stefan Priebe - Profihost AG ha scritto:
> Hi Paolo,
> Hi Josh,
>
> i started to port the iscsi fixes to rbd. The one think i'm missing is.
> How to tell / call rbd to cancel I/O on the server side?
>
> I grepped librbd source code for abort / cancel but didn't find anything.
Hi Paolo,
this is my current work status on porting these fixes to rbd. Right now
the discards get still canceled by the client kernel.
Might you have a look what i have forgotten?
Thanks!
Stefan
Am 19.11.2012 14:06, schrieb Paolo Bonzini:
Il 19/11/2012 14:01, Stefan Priebe - Profihost AG h
Hi Paolo,
Hi Josh,
i started to port the iscsi fixes to rbd. The one think i'm missing is.
How to tell / call rbd to cancel I/O on the server side?
I grepped librbd source code for abort / cancel but didn't find anything.
Qemu must be able to tell rbd to cancel a specific I/O request.
Greets
Il 19/11/2012 14:01, Stefan Priebe - Profihost AG ha scritto:
>> The right behavior is to return
>> only after the target says whether the cancellation was done or not.
>> For libiscsi, it was implemented by the commits you mention.
>
> So the whole bunch of changes is needed for rbd?
Something l
Am 19.11.2012 13:24, schrieb Paolo Bonzini:
Il 19/11/2012 12:49, Stefan Priebe - Profihost AG ha scritto:
It still causes hangs no? Though it works apart from that.
iscsi/libiscsi and discards works fine since your latest patches:
1bd075f29ea6d11853475c7c42734595720c3ac6
cfb3f5064af2d2e29c97
Il 19/11/2012 12:49, Stefan Priebe - Profihost AG ha scritto:
>>
>> It still causes hangs no? Though it works apart from that.
>
> iscsi/libiscsi and discards works fine since your latest patches:
> 1bd075f29ea6d11853475c7c42734595720c3ac6
> cfb3f5064af2d2e29c976e292c9472dfe9d61e31
> 27cbd828c617
Am 19.11.2012 12:16, schrieb Paolo Bonzini:
Il 19/11/2012 11:57, Stefan Priebe - Profihost AG ha scritto:
Yeah thats my old thread regarding iscsi und unmap but this works fine
now since you patched qemu.
It still causes hangs no? Though it works apart from that.
iscsi/libiscsi and discards
Il 19/11/2012 11:57, Stefan Priebe - Profihost AG ha scritto:
> Yeah thats my old thread regarding iscsi und unmap but this works fine
> now since you patched qemu.
It still causes hangs no? Though it works apart from that.
Paolo
> Stefan
>
> Am 19.11.2012 11:36, schrieb Paolo Bonzini:
>> Il 1
Yeah thats my old thread regarding iscsi und unmap but this works fine
now since you patched qemu.
Stefan
Am 19.11.2012 11:36, schrieb Paolo Bonzini:
Il 19/11/2012 11:30, Stefan Priebe - Profihost AG ha scritto:
But do you have any idea why it works with an iscsi / libiscsi backend?
In that
Il 19/11/2012 11:30, Stefan Priebe - Profihost AG ha scritto:
>
>
> But do you have any idea why it works with an iscsi / libiscsi backend?
> In that case the kernel does not cancel the I/O.
It tries to, but the command ultimately succeeds and the success
response "undoes" the cancel.
http://th
Am 19.11.2012 11:23, schrieb Paolo Bonzini:
Il 19/11/2012 11:13, Stefan Priebe - Profihost AG ha scritto:
So you mean this is not a bug in rbd or qemu this is a general bug in
the linux kernel since they implemented discard?
Yes.
As you're known in the linux dev community ;-) Might you op
Il 19/11/2012 11:13, Stefan Priebe - Profihost AG ha scritto:
>>>
>>> So you mean this is not a bug in rbd or qemu this is a general bug in
>>> the linux kernel since they implemented discard?
>>
>> Yes.
>
> As you're known in the linux dev community ;-) Might you open a thread /
> discussion
Am 19.11.2012 11:06, schrieb Paolo Bonzini:
Il 19/11/2012 10:59, Stefan Priebe - Profihost AG ha scritto:
Do you know what is the correct way?
I think the correct fix is to serialize them in the kernel.
So you mean this is not a bug in rbd or qemu this is a general bug in
the linux kernel
Il 19/11/2012 10:59, Stefan Priebe - Profihost AG ha scritto:
>>>
>>>
>>> Do you know what is the correct way?
>>
>> I think the correct fix is to serialize them in the kernel.
>
> So you mean this is not a bug in rbd or qemu this is a general bug in
> the linux kernel since they implemented disca
Am 19.11.2012 10:54, schrieb Paolo Bonzini:
Il 19/11/2012 10:36, Stefan Priebe - Profihost AG ha scritto:
Hi Paolo,
Am 19.11.2012 09:10, schrieb Paolo Bonzini:
I'm sorry the discard requests aren't failing. Qemu / Block driver
starts to cancel a bunch of requests.
That is being done in the k
Il 19/11/2012 10:36, Stefan Priebe - Profihost AG ha scritto:
> Hi Paolo,
>
> Am 19.11.2012 09:10, schrieb Paolo Bonzini:
>>> I'm sorry the discard requests aren't failing. Qemu / Block driver
>>> starts to cancel a bunch of requests.
>>
>> That is being done in the kernel (the guest, I think) bec
Hi Paolo,
Am 19.11.2012 09:10, schrieb Paolo Bonzini:
I'm sorry the discard requests aren't failing. Qemu / Block driver
starts to cancel a bunch of requests.
That is being done in the kernel (the guest, I think) because the UNMAPs
are taking too long.
That makes sense. RBD handles discards
Il 18/11/2012 23:00, Stefan Priebe ha scritto:
> Hi Paolo,
>
> Am 06.11.2012 23:42, schrieb Paolo Bonzini:
>>
>>> i wantes to use scsi unmap with rbd. rbd documention says you need to
>>> set discard_granularity=512 for the device. I'm using qemu 1.2.
>>>
>>> If i set this and send an UNMAP comman
Hi Paolo,
Am 06.11.2012 23:42, schrieb Paolo Bonzini:
i wantes to use scsi unmap with rbd. rbd documention says you need to
set discard_granularity=512 for the device. I'm using qemu 1.2.
If i set this and send an UNMAP command i get this kernel output:
The discard request is failing. Plea
Am 06.11.2012 23:42, schrieb Paolo Bonzini:
i wantes to use scsi unmap with rbd. rbd documention says you need to
set discard_granularity=512 for the device. I'm using qemu 1.2.
If i set this and send an UNMAP command i get this kernel output:
The discard request is failing. Please check wh
> i wantes to use scsi unmap with rbd. rbd documention says you need to
> set discard_granularity=512 for the device. I'm using qemu 1.2.
>
> If i set this and send an UNMAP command i get this kernel output:
The discard request is failing. Please check why with a breakpoint on
rbd_aio_discard_w
Hello list,
i wantes to use scsi unmap with rbd. rbd documention says you need to
set discard_granularity=512 for the device. I'm using qemu 1.2.
If i set this and send an UNMAP command i get this kernel output:
Sense Key : Aborted Command [current]
sd 2:0:0:1: [sdb]
Add. Sense: I/O process t
28 matches
Mail list logo
