From mboxrd@z Thu Jan 1 00:00:00 1970 From: "M. Mohan Kumar" Subject: Re: Default low water mark threshold for a new thin pool Date: Tue, 05 Feb 2013 22:26:15 +0530 Message-ID: <87boby1yz4.fsf@in.ibm.com> References: <87ehgv16gf.fsf@in.ibm.com> <20130205121846.GB3082@raspberrypi> Reply-To: LVM2 development Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20130205121846.GB3082@raspberrypi> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: lvm-devel-bounces@redhat.com Errors-To: lvm-devel-bounces@redhat.com To: thornber@redhat.com, LVM2 development Cc: dm-devel@redhat.com List-Id: dm-devel.ids thornber@redhat.com writes: > On Tue, Feb 05, 2013 at 02:30:00PM +0530, M. Mohan Kumar wrote: >> >> Hello, >> >> We are adding a lvm2api for creating a thin pool. During thin pool >> creation if user does not specify anything for the low water mark >> thresold, it would result in 0 and no events will be generated when thin >> pool capacity is reaching its full size. > > An event will be generated. Yeah, I missed that. But if thin pool is consumed fully there is no enough time to resize the pool and it may result in unwanted scenario? > >> I am not sure what should be the default low water threshold >> percentage. Is there any recommended value for low water threshold value >> in dm thin pools? > > It's very hard to say since it depends on your use case. In general > you want to allow dmeventd plenty of time to extend the pool before > you run out of space. So if you're seeing a _peak_ provisioning rate of > 100meg a minute, I'd set the lwm to about 100meg. It would be nice if LVM > could help you through this process. Do you mean mesaure how much MB/blocks dm-thin is able to provision in a minute and treat it as the default lwm blocks? > > - Joe > > -- > lvm-devel mailing list > lvm-devel@redhat.com > https://www.redhat.com/mailman/listinfo/lvm-devel -- lvm-devel mailing list lvm-devel@redhat.com https://www.redhat.com/mailman/listinfo/lvm-devel From mboxrd@z Thu Jan 1 00:00:00 1970 From: M. Mohan Kumar Date: Tue, 05 Feb 2013 22:26:15 +0530 Subject: Default low water mark threshold for a new thin pool In-Reply-To: <20130205121846.GB3082@raspberrypi> References: <87ehgv16gf.fsf@in.ibm.com> <20130205121846.GB3082@raspberrypi> Message-ID: <87boby1yz4.fsf@in.ibm.com> List-Id: To: lvm-devel@redhat.com MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit thornber at redhat.com writes: > On Tue, Feb 05, 2013 at 02:30:00PM +0530, M. Mohan Kumar wrote: >> >> Hello, >> >> We are adding a lvm2api for creating a thin pool. During thin pool >> creation if user does not specify anything for the low water mark >> thresold, it would result in 0 and no events will be generated when thin >> pool capacity is reaching its full size. > > An event will be generated. Yeah, I missed that. But if thin pool is consumed fully there is no enough time to resize the pool and it may result in unwanted scenario? > >> I am not sure what should be the default low water threshold >> percentage. Is there any recommended value for low water threshold value >> in dm thin pools? > > It's very hard to say since it depends on your use case. In general > you want to allow dmeventd plenty of time to extend the pool before > you run out of space. So if you're seeing a _peak_ provisioning rate of > 100meg a minute, I'd set the lwm to about 100meg. It would be nice if LVM > could help you through this process. Do you mean mesaure how much MB/blocks dm-thin is able to provision in a minute and treat it as the default lwm blocks? > > - Joe > > -- > lvm-devel mailing list > lvm-devel at redhat.com > https://www.redhat.com/mailman/listinfo/lvm-devel