Re: [lustre-discuss] ldiskfs ost size limit
Thanks Ben, is there a public document where I could have found this limit? Regards, Götz On Tue, Jul 21, 2015 at 4:33 PM, Ben Evans bev...@cray.com wrote: 128 TB is the current limit You can force more than that, but it looks like you won't need to. -Ben Evans -Original Message- From: lustre-discuss [mailto:lustre-discuss-boun...@lists.lustre.org] On Behalf Of Götz Waschk Sent: Tuesday, July 21, 2015 10:18 AM To: lustre-discuss@lists.lustre.org Subject: [lustre-discuss] ldiskfs ost size limit Dear Lustre experts, I'm in the process of installing a new Lustre file system based on version 2.5. What is the size limit for an OST when using ldiskfs?Can I format a 60 TB device with ldiskfs? Regards, Götz Waschk ___ lustre-discuss mailing list lustre-discuss@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org -- AL I:40: Do what thou wilt shall be the whole of the Law. ___ lustre-discuss mailing list lustre-discuss@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org
Re: [lustre-discuss] Speeding up recovery
1) You mention they are on the same host. Are they on separate partitions already? As you have failover configured I'm assuming that both servers can see the storage. In which case this will not be too difficult (depending on your failover software of course) if they have separate partitions. Yes, they are separate DRBD Devices. So mounting any one of them on the other server is easy. But how do I tell the OSS that MGS or MDT has moved to a new IP/Host? And how do I reconfigure the failover on the device I move? 2) so today Linux clients use the native client? And you are planning on shifting this to use the NFS service from a gateway node, is that correct? How do they connect to the lustre servers today? QDR IB? How will they reach the gateway nodes after this change? NFS over IB? NFS over RDMA? Yes, the Linux Hosts use Lustre Native Clients. Windows Hosts connect via the Gateway. The Gateway Nodes uses Infiniband+RDMA to connect to Lustre. I am thinking of moving the Linux Native Clients to NFS, connecting them through this Gateway. All client nodes are on 1GbE network. Infiniband is used only to connect the Gateway to Lustre. Regards, Indivar Nair On Tue, Jul 21, 2015 at 8:29 PM, Wahl, Edward ew...@osc.edu wrote: 1) You mention they are on the same host. Are they on separate partitions already? As you have failover configured I'm assuming that both servers can see the storage. In which case this will not be too difficult (depending on your failover software of course) if they have separate partitions. 2) so today Linux clients use the native client? And you are planning on shifting this to use the NFS service from a gateway node, is that correct? How do they connect to the lustre servers today? QDR IB? How will they reach the gateway nodes after this change? NFS over IB? NFS over RDMA? Ed -- *From:* lustre-discuss [lustre-discuss-boun...@lists.lustre.org] on behalf of Indivar Nair [indivar.n...@techterra.in] *Sent:* Tuesday, July 21, 2015 4:27 AM *To:* lustre-discuss; hpdd-discuss *Subject:* [lustre-discuss] Speeding up recovery Hi ..., Currently, Failover and Recovery takes a very long long time in our setup; almost 20 Minutes. We would like to make it as fast as possible. I have two queries regarding this - 1. === The MGS and MDT are on the same host. We do however have a passive stand-by server for the MGS/MDT server, which only mounts these partitions in case of a failure. *Current Setup* Server A: MGS+MDT Server B: Failover MGS+MDT I was wondering whether I can now move the MGS or MDT Partition to the standby server (so that imperative recovery works properly) - *New Setup* Server A: MDT *Failover MGS* Server B: *MGS* Failover MDT *OR * Server A: *MGS* Failover MDT Server B: MDT *Failover MGS* i.e. *Can I separate the MDT and MGS partitions on to different machines without formatting or reinstalling Lustre? * === 2. === This storage is used by around 150 Workstations and 150 Compute (Render) Nodes. Out of these 150 workstations, around 30 - 40 are MS Windows. The MS Windows clients access the storage through a 2-node Samba Gateway Cluster. The Gateway Nodes are connected to the storage through a QDR Infiniband Network. We were thinking of adding NFS Service to the Samba Gateway nodes, and reconfiguring the Linux clients to connect via this gateway. This will bring down the direct Lustre Clients to just 2 nodes. *So, will having only 2 clients improve the failover-recovery time?* === Is there anything else we can do to speed up recovery? Regards, Indivar Nair ___ lustre-discuss mailing list lustre-discuss@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org
Re: [lustre-discuss] Speeding up recovery
1) You mention they are on the same host. Are they on separate partitions already? As you have failover configured I'm assuming that both servers can see the storage. In which case this will not be too difficult (depending on your failover software of course) if they have separate partitions. 2) so today Linux clients use the native client? And you are planning on shifting this to use the NFS service from a gateway node, is that correct? How do they connect to the lustre servers today? QDR IB? How will they reach the gateway nodes after this change? NFS over IB? NFS over RDMA? Ed From: lustre-discuss [lustre-discuss-boun...@lists.lustre.org] on behalf of Indivar Nair [indivar.n...@techterra.in] Sent: Tuesday, July 21, 2015 4:27 AM To: lustre-discuss; hpdd-discuss Subject: [lustre-discuss] Speeding up recovery Hi ..., Currently, Failover and Recovery takes a very long long time in our setup; almost 20 Minutes. We would like to make it as fast as possible. I have two queries regarding this - 1. === The MGS and MDT are on the same host. We do however have a passive stand-by server for the MGS/MDT server, which only mounts these partitions in case of a failure. Current Setup Server A: MGS+MDT Server B: Failover MGS+MDT I was wondering whether I can now move the MGS or MDT Partition to the standby server (so that imperative recovery works properly) - New Setup Server A: MDT Failover MGS Server B: MGS Failover MDT OR Server A: MGS Failover MDT Server B: MDT Failover MGS i.e. Can I separate the MDT and MGS partitions on to different machines without formatting or reinstalling Lustre? === 2. === This storage is used by around 150 Workstations and 150 Compute (Render) Nodes. Out of these 150 workstations, around 30 - 40 are MS Windows. The MS Windows clients access the storage through a 2-node Samba Gateway Cluster. The Gateway Nodes are connected to the storage through a QDR Infiniband Network. We were thinking of adding NFS Service to the Samba Gateway nodes, and reconfiguring the Linux clients to connect via this gateway. This will bring down the direct Lustre Clients to just 2 nodes. So, will having only 2 clients improve the failover-recovery time? === Is there anything else we can do to speed up recovery? Regards, Indivar Nair ___ lustre-discuss mailing list lustre-discuss@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org
[lustre-discuss] Lustre Server Sizing
Hi ..., One of our customers has a 3 x 240 Disk SAN Storage Array and would like to convert it to Lustre. They have around 150 Workstations and around 200 Compute (Render) nodes. The File Sizes they generally work with are - 1 to 1.5 million files (images) of 10-20MB in size. And a few thousand files of 500-1000MB in size. Almost 50% of the infra is on MS Windows or Apple MACs I was thinking of the following configuration - 1 MDS 1 Failover MDS 3 OSS (failover to each other) 3 NFS+CIFS Gateway Servers FDR Infiniband backend network (to connect the Gateways to Lustre) Each Gateway Server will have 8 x 10GbE Frontend Network (connecting the clients) *Option A* 10+10 Disk RAID60 Array with 64KB Chunk Size i.e. 1MB Stripe Width 720 Disks / (10+10) = 36 Arrays. 12 OSTs per OSS 18 OSTs per OSS in case of Failover *Option B* 10+10+10+10 Disk RAID60 Array with 128KB Chunk Size i.e. 4MB Stripe Width 720 Disks / (10+10+10+10) = 18 Arrays 6 OSTs per OSS 9 OSTs per OSS in case of Failover 4MB RPC and I/O *Questions* 1. Would it be better to let Lustre do most of the striping / file distribution (as in Option A) OR would it be better to let the RAID Controllers do it (as in Option B) 2. Will Option B allow us to have lesser CPU/RAM than Option A? Regards, Indivar Nair ___ lustre-discuss mailing list lustre-discuss@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org
Re: [lustre-discuss] A quick question about reusing osts (lustre 2.5.3)
I was wondering because while looking at the man page for mkfs.lustre I saw the below option: --replace Used to initialize a target with the same --index as a previously used target if the old target was permanently lost for some reason (e.g. multiple disk failure or massive corruption). This avoids having the target try to register as a new target with the MGS. w/r, Kurt - Original Message - From: Ben Evans bev...@cray.com To: Kurt Strosahl stros...@jlab.org, lustre-discuss@lists.lustre.org Sent: Tuesday, July 21, 2015 10:17:47 AM Subject: RE: A quick question about reusing osts (lustre 2.5.3) I know you'd need to keep the config files, directory structures, etc. How much of that info you need to keep around, I'm not 100% sure. To get the MGS to accept it again, you may have to unmount and run tunefs.lustre --writeconf on all the targets. -Ben Evans -Original Message- From: lustre-discuss [mailto:lustre-discuss-boun...@lists.lustre.org] On Behalf Of Kurt Strosahl Sent: Tuesday, July 21, 2015 9:15 AM To: lustre-discuss@lists.lustre.org Subject: [lustre-discuss] A quick question about reusing osts (lustre 2.5.3) Hello, I had a quick question about recreating osts... If I drain all the files off an ost can I just reformat it and have it added back into lustre, in essence reusing the same index? The server wouldn't change. Or would I have to preserve its configuration files? w/r, Kurt ___ lustre-discuss mailing list lustre-discuss@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org ___ lustre-discuss mailing list lustre-discuss@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org
Re: [lustre-discuss] ldiskfs ost size limit
It is available in the Lustre Operations Manual, Table 1.1 https://build.hpdd.intel.com/job/lustre-manual/lastSuccessfulBuild/artifact/lustre_manual.xhtml#idp162992 - Justin On 7/21/15, 11:02 AM, lustre-discuss on behalf of Götz Waschk lustre-discuss-boun...@lists.lustre.org on behalf of goetz.was...@gmail.com wrote: Thanks Ben, is there a public document where I could have found this limit? Regards, Götz On Tue, Jul 21, 2015 at 4:33 PM, Ben Evans bev...@cray.com wrote: 128 TB is the current limit You can force more than that, but it looks like you won't need to. -Ben Evans -Original Message- From: lustre-discuss [mailto:lustre-discuss-boun...@lists.lustre.org] On Behalf Of Götz Waschk Sent: Tuesday, July 21, 2015 10:18 AM To: lustre-discuss@lists.lustre.org Subject: [lustre-discuss] ldiskfs ost size limit Dear Lustre experts, I'm in the process of installing a new Lustre file system based on version 2.5. What is the size limit for an OST when using ldiskfs?Can I format a 60 TB device with ldiskfs? Regards, Götz Waschk ___ lustre-discuss mailing list lustre-discuss@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org -- AL I:40: Do what thou wilt shall be the whole of the Law. ___ lustre-discuss mailing list lustre-discuss@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org ___ lustre-discuss mailing list lustre-discuss@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org
Re: [lustre-discuss] [HPDD-discuss] Lustre Server Sizing
Hi Scott, The 3 - SAN Storages with 240 disks each has its own 3 NAS Headers (NAS Appliances). However, even with 240 10K RPM disk and RAID50, it is only providing around 1.2 - 1.4GB/s per NAS Header. There is no clustered file system, and each NAS Header has its own file-system. It uses some custom mechanism to present the 3 file systems as single name space. But the directories have to be manually spread across for load-balancing. As you can guess, this doesn't work most of the time. Many a times, most of the compute nodes access a single NAS Header, overloading it. The customer wants *at least* 9GB/s throughput from a single file-system. But I think, if we architect the Lustre Storage correctly, with these many disks, we should get at least 18GB/s throughput, if not more. Regards, Indivar Nair On Tue, Jul 21, 2015 at 10:15 PM, Scott Nolin scott.no...@ssec.wisc.edu wrote: An important question is what performance do they have now, and what do they expect if converting it to Lustre. Our more basically, what are they looking for in general in changing? The performance requirements may help drive your OSS numbers for example, or interconnect, and all kinds of stuff. Also I don't have a lot of experience with NFS/CIFS gateways, but that is perhaps it's own topic and may need some close attention. Scott On 7/21/2015 10:57 AM, Indivar Nair wrote: Hi ..., One of our customers has a 3 x 240 Disk SAN Storage Array and would like to convert it to Lustre. They have around 150 Workstations and around 200 Compute (Render) nodes. The File Sizes they generally work with are - 1 to 1.5 million files (images) of 10-20MB in size. And a few thousand files of 500-1000MB in size. Almost 50% of the infra is on MS Windows or Apple MACs I was thinking of the following configuration - 1 MDS 1 Failover MDS 3 OSS (failover to each other) 3 NFS+CIFS Gateway Servers FDR Infiniband backend network (to connect the Gateways to Lustre) Each Gateway Server will have 8 x 10GbE Frontend Network (connecting the clients) *Option A* 10+10 Disk RAID60 Array with 64KB Chunk Size i.e. 1MB Stripe Width 720 Disks / (10+10) = 36 Arrays. 12 OSTs per OSS 18 OSTs per OSS in case of Failover *Option B* 10+10+10+10 Disk RAID60 Array with 128KB Chunk Size i.e. 4MB Stripe Width 720 Disks / (10+10+10+10) = 18 Arrays 6 OSTs per OSS 9 OSTs per OSS in case of Failover 4MB RPC and I/O *Questions* 1. Would it be better to let Lustre do most of the striping / file distribution (as in Option A) OR would it be better to let the RAID Controllers do it (as in Option B) 2. Will Option B allow us to have lesser CPU/RAM than Option A? Regards, Indivar Nair ___ HPDD-discuss mailing list hpdd-disc...@lists.01.org https://lists.01.org/mailman/listinfo/hpdd-discuss ___ lustre-discuss mailing list lustre-discuss@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org ___ lustre-discuss mailing list lustre-discuss@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org
Re: [lustre-discuss] Lustre Server Sizing
Note the other email also seemed to suggest that multiple NFS exports of Lustre wouldn't work. I don't think that's the case, as we have this sort of setup at a number of our customers without particular trouble. In the abstract, I could see the possibility of some caching errors between different clients, but that would be only namespace stuff, not data. And I think in practice that's ok. But regardless, as Andreas said, for the Linux clients, Lustre directly will give much better results. From: lustre-discuss [lustre-discuss-boun...@lists.lustre.org] on behalf of Dilger, Andreas [andreas.dil...@intel.com] Sent: Tuesday, July 21, 2015 6:59 PM To: Indivar Nair Cc: hpdd-discuss; to: lustre-discuss Subject: Re: [lustre-discuss] Lustre Server Sizing Having only 3 OSS will limit the performance you can get, and having so many OSTs on each OSS will give sub-optimal performance. 4-6 OSTs/OSS is more reasonable. It also isn't clear why you want RAID-60 instead of just RAID-10? Finally, for Linux clients it is much better to use direct Lustre access instead of NFS as mentioned in another email. Cheers, Andreas On Jul 21, 2015, at 08:58, Indivar Nair indivar.n...@techterra.inmailto:indivar.n...@techterra.in wrote: Hi ..., One of our customers has a 3 x 240 Disk SAN Storage Array and would like to convert it to Lustre. They have around 150 Workstations and around 200 Compute (Render) nodes. The File Sizes they generally work with are - 1 to 1.5 million files (images) of 10-20MB in size. And a few thousand files of 500-1000MB in size. Almost 50% of the infra is on MS Windows or Apple MACs I was thinking of the following configuration - 1 MDS 1 Failover MDS 3 OSS (failover to each other) 3 NFS+CIFS Gateway Servers FDR Infiniband backend network (to connect the Gateways to Lustre) Each Gateway Server will have 8 x 10GbE Frontend Network (connecting the clients) Option A 10+10 Disk RAID60 Array with 64KB Chunk Size i.e. 1MB Stripe Width 720 Disks / (10+10) = 36 Arrays. 12 OSTs per OSS 18 OSTs per OSS in case of Failover Option B 10+10+10+10 Disk RAID60 Array with 128KB Chunk Size i.e. 4MB Stripe Width 720 Disks / (10+10+10+10) = 18 Arrays 6 OSTs per OSS 9 OSTs per OSS in case of Failover 4MB RPC and I/O Questions 1. Would it be better to let Lustre do most of the striping / file distribution (as in Option A) OR would it be better to let the RAID Controllers do it (as in Option B) 2. Will Option B allow us to have lesser CPU/RAM than Option A? Regards, Indivar Nair ___ lustre-discuss mailing list lustre-discuss@lists.lustre.orgmailto:lustre-discuss@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org ___ lustre-discuss mailing list lustre-discuss@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org ___ lustre-discuss mailing list lustre-discuss@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org
Re: [lustre-discuss] [HPDD-discuss] Lustre Server Sizing
I’ve seen CTDB + Samba deployed on several sites running Lustre. It’s stable in my experience, and straightforward to get installed and set up, although the process is time-consuming. The most significant hurdle is integrating with AD and maybe load balancing for the CTDB servers (RR DNS is the easiest and most common solution). Performance is not nearly as good as for native Lustre client (apart from anything else, IIRC, SMB is a “chatty” protocol, esp with xattrs?). One downside of CTDB is that Lustre client must be mounted with -oflock in order for the recovery lock manager to work. Each individual connection to Samba from a Windows client is limited to the bandwidth and single thread performance of the CTDB node. Clients remain connected to a single CTDB node for the duration of their session, so there is a possibility of an imbalance in connections over time. Load balancing is strictly round-robin through DNS lookups, unless a more sophisticated load balancer is placed in front of the CTDB cluster. There are references to CTDB + NFS / Ganesha as well but I haven’t had an opportunity to try it out. Most of the demand for non-native client access to Lustre involves Windows machines. Malcolm. From: lustre-discuss [mailto:lustre-discuss-boun...@lists.lustre.org] On Behalf Of Jeff Johnson Sent: Wednesday, July 22, 2015 5:54 AM To: Indivar Nair Cc: lustre-discuss Subject: Re: [lustre-discuss] [HPDD-discuss] Lustre Server Sizing Indivar, Since your CIFS or NFS gateways operate as Lustre clients there can be issues with running multiple NFS or CIFS gateway machines frontending the same Lustre filesystem. As Lustre clients there are no issues in terms of file locking but the NFS and CIFS caching and multi-client file access mechanics don't interface with Lustre's file locking mechanics. Perhaps that may have changed recently and a developer on the list may comment on developments there. So while you could provide client access through multiple NFS or CIFS gateway machines there would not be much in the way of file locking protection. There is a way to configure pCIFS with CTDB and get close to what you envision with Samba. I did that configuration once as a proof of concept (no valuable data). It is a *very* complex configuration and based on the state of software when I did it I wouldn't say it was a production grade environment. As I said before, my understanding may be a year out of date and someone else could speak to the current state of things. Hopefully that would be a better story. --Jeff On Tue, Jul 21, 2015 at 10:26 AM, Indivar Nair indivar.n...@techterra.inmailto:indivar.n...@techterra.in wrote: Hi Scott, The 3 - SAN Storages with 240 disks each has its own 3 NAS Headers (NAS Appliances). However, even with 240 10K RPM disk and RAID50, it is only providing around 1.2 - 1.4GB/s per NAS Header. There is no clustered file system, and each NAS Header has its own file-system. It uses some custom mechanism to present the 3 file systems as single name space. But the directories have to be manually spread across for load-balancing. As you can guess, this doesn't work most of the time. Many a times, most of the compute nodes access a single NAS Header, overloading it. The customer wants *at least* 9GB/s throughput from a single file-system. But I think, if we architect the Lustre Storage correctly, with these many disks, we should get at least 18GB/s throughput, if not more. Regards, Indivar Nair On Tue, Jul 21, 2015 at 10:15 PM, Scott Nolin scott.no...@ssec.wisc.edumailto:scott.no...@ssec.wisc.edu wrote: An important question is what performance do they have now, and what do they expect if converting it to Lustre. Our more basically, what are they looking for in general in changing? The performance requirements may help drive your OSS numbers for example, or interconnect, and all kinds of stuff. Also I don't have a lot of experience with NFS/CIFS gateways, but that is perhaps it's own topic and may need some close attention. Scott On 7/21/2015 10:57 AM, Indivar Nair wrote: Hi ..., One of our customers has a 3 x 240 Disk SAN Storage Array and would like to convert it to Lustre. They have around 150 Workstations and around 200 Compute (Render) nodes. The File Sizes they generally work with are - 1 to 1.5 million files (images) of 10-20MB in size. And a few thousand files of 500-1000MB in size. Almost 50% of the infra is on MS Windows or Apple MACs I was thinking of the following configuration - 1 MDS 1 Failover MDS 3 OSS (failover to each other) 3 NFS+CIFS Gateway Servers FDR Infiniband backend network (to connect the Gateways to Lustre) Each Gateway Server will have 8 x 10GbE Frontend Network (connecting the clients) *Option A* 10+10 Disk RAID60 Array with 64KB Chunk Size i.e. 1MB Stripe Width 720 Disks / (10+10) = 36 Arrays. 12 OSTs per OSS 18 OSTs per OSS in case of Failover *Option B* 10+10+10+10 Disk RAID60
Re: [lustre-discuss] [HPDD-discuss] Lustre Server Sizing
Indivar, Since your CIFS or NFS gateways operate as Lustre clients there can be issues with running multiple NFS or CIFS gateway machines frontending the same Lustre filesystem. As Lustre clients there are no issues in terms of file locking but the NFS and CIFS caching and multi-client file access mechanics don't interface with Lustre's file locking mechanics. Perhaps that may have changed recently and a developer on the list may comment on developments there. So while you could provide client access through multiple NFS or CIFS gateway machines there would not be much in the way of file locking protection. There is a way to configure pCIFS with CTDB and get close to what you envision with Samba. I did that configuration once as a proof of concept (no valuable data). It is a *very* complex configuration and based on the state of software when I did it I wouldn't say it was a production grade environment. As I said before, my understanding may be a year out of date and someone else could speak to the current state of things. Hopefully that would be a better story. --Jeff On Tue, Jul 21, 2015 at 10:26 AM, Indivar Nair indivar.n...@techterra.in wrote: Hi Scott, The 3 - SAN Storages with 240 disks each has its own 3 NAS Headers (NAS Appliances). However, even with 240 10K RPM disk and RAID50, it is only providing around 1.2 - 1.4GB/s per NAS Header. There is no clustered file system, and each NAS Header has its own file-system. It uses some custom mechanism to present the 3 file systems as single name space. But the directories have to be manually spread across for load-balancing. As you can guess, this doesn't work most of the time. Many a times, most of the compute nodes access a single NAS Header, overloading it. The customer wants *at least* 9GB/s throughput from a single file-system. But I think, if we architect the Lustre Storage correctly, with these many disks, we should get at least 18GB/s throughput, if not more. Regards, Indivar Nair On Tue, Jul 21, 2015 at 10:15 PM, Scott Nolin scott.no...@ssec.wisc.edu wrote: An important question is what performance do they have now, and what do they expect if converting it to Lustre. Our more basically, what are they looking for in general in changing? The performance requirements may help drive your OSS numbers for example, or interconnect, and all kinds of stuff. Also I don't have a lot of experience with NFS/CIFS gateways, but that is perhaps it's own topic and may need some close attention. Scott On 7/21/2015 10:57 AM, Indivar Nair wrote: Hi ..., One of our customers has a 3 x 240 Disk SAN Storage Array and would like to convert it to Lustre. They have around 150 Workstations and around 200 Compute (Render) nodes. The File Sizes they generally work with are - 1 to 1.5 million files (images) of 10-20MB in size. And a few thousand files of 500-1000MB in size. Almost 50% of the infra is on MS Windows or Apple MACs I was thinking of the following configuration - 1 MDS 1 Failover MDS 3 OSS (failover to each other) 3 NFS+CIFS Gateway Servers FDR Infiniband backend network (to connect the Gateways to Lustre) Each Gateway Server will have 8 x 10GbE Frontend Network (connecting the clients) *Option A* 10+10 Disk RAID60 Array with 64KB Chunk Size i.e. 1MB Stripe Width 720 Disks / (10+10) = 36 Arrays. 12 OSTs per OSS 18 OSTs per OSS in case of Failover *Option B* 10+10+10+10 Disk RAID60 Array with 128KB Chunk Size i.e. 4MB Stripe Width 720 Disks / (10+10+10+10) = 18 Arrays 6 OSTs per OSS 9 OSTs per OSS in case of Failover 4MB RPC and I/O *Questions* 1. Would it be better to let Lustre do most of the striping / file distribution (as in Option A) OR would it be better to let the RAID Controllers do it (as in Option B) 2. Will Option B allow us to have lesser CPU/RAM than Option A? Regards, Indivar Nair ___ HPDD-discuss mailing list hpdd-disc...@lists.01.org https://lists.01.org/mailman/listinfo/hpdd-discuss ___ lustre-discuss mailing list lustre-discuss@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org ___ lustre-discuss mailing list lustre-discuss@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org -- -- Jeff Johnson Co-Founder Aeon Computing jeff.john...@aeoncomputing.com www.aeoncomputing.com t: 858-412-3810 x1001 f: 858-412-3845 m: 619-204-9061 4170 Morena Boulevard, Suite D - San Diego, CA 92117 High-Performance Computing / Lustre Filesystems / Scale-out Storage ___ lustre-discuss mailing list lustre-discuss@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org
Re: [lustre-discuss] How Lustre stores hyperslabs and chunks of HDF5?
Currently there is no direct connection between Lustre layout and HDF5 file layout. The only option is RAID-0 striping across OST objects with a fixed stripe size. If HDF5 is aware of this stripe size and can take advantage of it, that is great. There is a project that has started to implement Progressive File Layout (PFL) that allows different extents of a file to have different stripe counts and stripe sizes, which could be leveraged by libraries like HDF5 in the future. See http://cdn.opensfs.org/wp-content/uploads/2015/04/Progressive-File-Layouts_Hammond.pdf and/or https://www.youtube.com/watch?v=5rm6Nlmqdp0 for more details on PFL prototype development. Cheers, Andreas On Jul 19, 2015, at 20:37, prakrati.agra...@shell.commailto:prakrati.agra...@shell.com prakrati.agra...@shell.commailto:prakrati.agra...@shell.com wrote: Hi, I wanted to understand how Lustre stores the chunks and hyperslabs in HDF5 framework on the OSTs? If I set the chunk size and each rank is writing a hyperslab, then OST0 has chunk0, OST1 has chunk1 and so on or is it that OST0 has hyperslab0, OST1 has hyperslab1. Is there any way of finding that out? Thanks and Regards, Prakrati ___ lustre-discuss mailing list lustre-discuss@lists.lustre.orgmailto:lustre-discuss@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org ___ lustre-discuss mailing list lustre-discuss@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org
Re: [lustre-discuss] Speeding up recovery
I believe this is described in the Lustre Manual, but the basic process to split a combined MDS+MGS into a separate MGS is to format a new MGS device, then copy all the files from CONFIGS on the old combined MDT+MGT device into the new MGS. See the manual for full details. Cheers, Andreas On Jul 21, 2015, at 01:27, Indivar Nair indivar.n...@techterra.inmailto:indivar.n...@techterra.in wrote: Hi ..., Currently, Failover and Recovery takes a very long long time in our setup; almost 20 Minutes. We would like to make it as fast as possible. I have two queries regarding this - 1. === The MGS and MDT are on the same host. We do however have a passive stand-by server for the MGS/MDT server, which only mounts these partitions in case of a failure. Current Setup Server A: MGS+MDT Server B: Failover MGS+MDT I was wondering whether I can now move the MGS or MDT Partition to the standby server (so that imperative recovery works properly) - New Setup Server A: MDT Failover MGS Server B: MGS Failover MDT OR Server A: MGS Failover MDT Server B: MDT Failover MGS i.e. Can I separate the MDT and MGS partitions on to different machines without formatting or reinstalling Lustre? === 2. === This storage is used by around 150 Workstations and 150 Compute (Render) Nodes. Out of these 150 workstations, around 30 - 40 are MS Windows. The MS Windows clients access the storage through a 2-node Samba Gateway Cluster. The Gateway Nodes are connected to the storage through a QDR Infiniband Network. We were thinking of adding NFS Service to the Samba Gateway nodes, and reconfiguring the Linux clients to connect via this gateway. This will bring down the direct Lustre Clients to just 2 nodes. So, will having only 2 clients improve the failover-recovery time? === Is there anything else we can do to speed up recovery? Regards, Indivar Nair ___ lustre-discuss mailing list lustre-discuss@lists.lustre.orgmailto:lustre-discuss@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org ___ lustre-discuss mailing list lustre-discuss@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org
Re: [lustre-discuss] Lustre Server Sizing
Having only 3 OSS will limit the performance you can get, and having so many OSTs on each OSS will give sub-optimal performance. 4-6 OSTs/OSS is more reasonable. It also isn't clear why you want RAID-60 instead of just RAID-10? Finally, for Linux clients it is much better to use direct Lustre access instead of NFS as mentioned in another email. Cheers, Andreas On Jul 21, 2015, at 08:58, Indivar Nair indivar.n...@techterra.inmailto:indivar.n...@techterra.in wrote: Hi ..., One of our customers has a 3 x 240 Disk SAN Storage Array and would like to convert it to Lustre. They have around 150 Workstations and around 200 Compute (Render) nodes. The File Sizes they generally work with are - 1 to 1.5 million files (images) of 10-20MB in size. And a few thousand files of 500-1000MB in size. Almost 50% of the infra is on MS Windows or Apple MACs I was thinking of the following configuration - 1 MDS 1 Failover MDS 3 OSS (failover to each other) 3 NFS+CIFS Gateway Servers FDR Infiniband backend network (to connect the Gateways to Lustre) Each Gateway Server will have 8 x 10GbE Frontend Network (connecting the clients) Option A 10+10 Disk RAID60 Array with 64KB Chunk Size i.e. 1MB Stripe Width 720 Disks / (10+10) = 36 Arrays. 12 OSTs per OSS 18 OSTs per OSS in case of Failover Option B 10+10+10+10 Disk RAID60 Array with 128KB Chunk Size i.e. 4MB Stripe Width 720 Disks / (10+10+10+10) = 18 Arrays 6 OSTs per OSS 9 OSTs per OSS in case of Failover 4MB RPC and I/O Questions 1. Would it be better to let Lustre do most of the striping / file distribution (as in Option A) OR would it be better to let the RAID Controllers do it (as in Option B) 2. Will Option B allow us to have lesser CPU/RAM than Option A? Regards, Indivar Nair ___ lustre-discuss mailing list lustre-discuss@lists.lustre.orgmailto:lustre-discuss@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org ___ lustre-discuss mailing list lustre-discuss@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org
[lustre-discuss] Speeding up recovery
Hi ..., Currently, Failover and Recovery takes a very long long time in our setup; almost 20 Minutes. We would like to make it as fast as possible. I have two queries regarding this - 1. === The MGS and MDT are on the same host. We do however have a passive stand-by server for the MGS/MDT server, which only mounts these partitions in case of a failure. *Current Setup* Server A: MGS+MDT Server B: Failover MGS+MDT I was wondering whether I can now move the MGS or MDT Partition to the standby server (so that imperative recovery works properly) - *New Setup* Server A: MDT *Failover MGS* Server B: *MGS* Failover MDT *OR* Server A: *MGS* Failover MDT Server B: MDT *Failover MGS* i.e. *Can I separate the MDT and MGS partitions on to different machines without formatting or reinstalling Lustre?* === 2. === This storage is used by around 150 Workstations and 150 Compute (Render) Nodes. Out of these 150 workstations, around 30 - 40 are MS Windows. The MS Windows clients access the storage through a 2-node Samba Gateway Cluster. The Gateway Nodes are connected to the storage through a QDR Infiniband Network. We were thinking of adding NFS Service to the Samba Gateway nodes, and reconfiguring the Linux clients to connect via this gateway. This will bring down the direct Lustre Clients to just 2 nodes. *So, will having only 2 clients improve the failover-recovery time?* === Is there anything else we can do to speed up recovery? Regards, Indivar Nair ___ lustre-discuss mailing list lustre-discuss@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org
[lustre-discuss] A quick question about reusing osts (lustre 2.5.3)
Hello, I had a quick question about recreating osts... If I drain all the files off an ost can I just reformat it and have it added back into lustre, in essence reusing the same index? The server wouldn't change. Or would I have to preserve its configuration files? w/r, Kurt ___ lustre-discuss mailing list lustre-discuss@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org
Re: [lustre-discuss] A quick question about reusing osts (lustre 2.5.3)
I know you'd need to keep the config files, directory structures, etc. How much of that info you need to keep around, I'm not 100% sure. To get the MGS to accept it again, you may have to unmount and run tunefs.lustre --writeconf on all the targets. -Ben Evans -Original Message- From: lustre-discuss [mailto:lustre-discuss-boun...@lists.lustre.org] On Behalf Of Kurt Strosahl Sent: Tuesday, July 21, 2015 9:15 AM To: lustre-discuss@lists.lustre.org Subject: [lustre-discuss] A quick question about reusing osts (lustre 2.5.3) Hello, I had a quick question about recreating osts... If I drain all the files off an ost can I just reformat it and have it added back into lustre, in essence reusing the same index? The server wouldn't change. Or would I have to preserve its configuration files? w/r, Kurt ___ lustre-discuss mailing list lustre-discuss@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org ___ lustre-discuss mailing list lustre-discuss@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org
[lustre-discuss] ldiskfs ost size limit
Dear Lustre experts, I'm in the process of installing a new Lustre file system based on version 2.5. What is the size limit for an OST when using ldiskfs?Can I format a 60 TB device with ldiskfs? Regards, Götz Waschk ___ lustre-discuss mailing list lustre-discuss@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org
Re: [lustre-discuss] ldiskfs ost size limit
128 TB is the current limit You can force more than that, but it looks like you won't need to. -Ben Evans -Original Message- From: lustre-discuss [mailto:lustre-discuss-boun...@lists.lustre.org] On Behalf Of Götz Waschk Sent: Tuesday, July 21, 2015 10:18 AM To: lustre-discuss@lists.lustre.org Subject: [lustre-discuss] ldiskfs ost size limit Dear Lustre experts, I'm in the process of installing a new Lustre file system based on version 2.5. What is the size limit for an OST when using ldiskfs?Can I format a 60 TB device with ldiskfs? Regards, Götz Waschk ___ lustre-discuss mailing list lustre-discuss@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org ___ lustre-discuss mailing list lustre-discuss@lists.lustre.org http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org