Re: TSM 5.5 using LANFree backup really slow
Flavio Is the destination of your backup pool tape? Have you confirmed with VALIDATE LANFREE on the server Are you seeing a proxied by message in the client log to indicate that the storage agent is in use? Is the *network* path configured as 1Gb - I'm suspecting you have a LAN backup and that the lan path has issues. HTH Regards Steve Steven Harris. TSM Admin, looking for work, Sydney Australia Flavio Junior wrote: Hi folks, I'm here, again, asking for help :) More I read, more doubts I have. I'm now implementing a TSM For SAN setup, to performance comparison with my currently LAN backup. I'll describe my scenario below, any doubt's feel free to ask me: - 2 Production nodes using RHEL 5.3 connected to SAN - 1 IBM DS4700 storage with 14 FC disks (13 raid5 + 1 hot-spare) - Each node have a emulex dual-port HBA - 2 Brocade TotalStorage fiber switches, with coherent zonings - Library is a IBM TS3100 LTO4 SAN, with only one drive Ok, i'm using TSM with LAN backups for a week, the transfer rate comes among 60~80GB per hour. The tape has a nominal write speed of 120MB/s that says me 430GB/h, ok.. this is a *nominal* speed, I was expecting something near 300GB/h, good enough for me. For these tests I've configured TSM Server on node1 and TSM Client on node2. Each node has 6 1Gbps NIC's, 4 to public network 2 for cluster-heartbeat + backups. With this setup I've got, as said above, 60~80GB/h. So, I decide to setup TSM for SAN.. Download a lot of docs/books and start reading/doing. I end up with: - Fabric zoning changes, isolating a HBA port on each dual-port cards for nodes - Installed StorageAgent on node2 - Installed TSM Client on node2 - - Configured the client to connect on storageagent - Redefine all my TSM Server setup - - Make Library as shared=yes, configure storage agent as a server, configure path, configure datawritepath=lanfree and so on... Ok, time to backup. And, for my surprise i got only ~500MB per minute.. I thought it very strange.. I've no idea what is the problem, I've tried with compression and without, DATAREADPATH=LAN and SAN, VirtualMountPoints or Physical ... Does anybody have an idea the reason of that bad transfer rate?? The servers are Samba Servers, with all kind of file and all size As I said before, I'm really new with TSM so is really possible that I've misconfigured something :) Anyway, thanks in advance, and any help will be appreciated. -- Flávio do Carmo Júnior aka waKKu Florianópolis/SC - Brazil No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.38/2274 - Release Date: 07/31/09 05:58:00
Re: Thumbs up/down for a new install for TSM 6.1 on Windows?
On 31/07, Wanda Prather wrote: Looking for opinions... I'm about to install a TSM server for a new customer who has never had TSM before. Very standard TSM config - - Plenty powerful Windows host - 100 mostly windows clients, back up to diskpool, migrate to 3592 tape drives, copy to 3592 tape and vault. - 2 TB of disk for diskpool, DB, logs if I need it. - No need for backupsets. I know there are still folks reporting problems with the TSM for Linux server, and conversion of existing 5.X DB's is non-trivial. But I haven't seen anything in the TSM flashes open APARS to think I'm taking much of a risk installing TSM 6.1 + patch on Windows. And it will save them the grief of doing the upgrade a year from now. What does the ADSM-L world think? Thanks for any input W My rule of thumb for going with new versions of (IBM) software is 6-12 months after release. Before summer I told customers to go with 5.5, now after the summer I'm going to go with 6.1. I dont feel totally comfortable with 6.1 yet but you gotta take the plunge sometime :) Especially since the requirements for DB storage is so different for 6.1 and because of the awkward upgrade process I believe the arguments for 6.1 are pretty strong. -km
Re: TSM 5.5 using LANFree backup really slow
Hi Steve, thanks for answering... Yeah, the destination is tape, i see on StorageAgent console the connection to server and opening tape, during the backup I see the bytes increasing on Backup by LANFree: (but, this value is ALWAYS lower than the value of normal backup... i don't know if this is normal). VALIDATE LANFREE node policyset shows me that the node is capable to LANFree backups. During the backup (using webclient interface) I can see two speeds, well.. A screenshot of backup report is at link: http://img18.imageshack.us/img18/9278/backupz.jpg The screenshot is in portuguese but I think is easy to you guess what is each value... Time: 5:45 Hours Size: 181 Gb No compression Rede == Network Agregada == Aggregate, combination... (not sure how TSM show it) Ok, this is what I have... If need more info just let me know. Thanks again. -- Flávio do Carmo Júnior aka waKKu Florianópolis/SC - Brazil On Sat, Aug 1, 2009 at 3:51 AM, Steven Harrisst...@stevenharris.info wrote: Flavio Is the destination of your backup pool tape? Have you confirmed with VALIDATE LANFREE on the server Are you seeing a proxied by message in the client log to indicate that the storage agent is in use? Is the *network* path configured as 1Gb - I'm suspecting you have a LAN backup and that the lan path has issues. HTH Regards Steve Steven Harris. TSM Admin, looking for work, Sydney Australia Flavio Junior wrote: Hi folks, I'm here, again, asking for help :) More I read, more doubts I have. I'm now implementing a TSM For SAN setup, to performance comparison with my currently LAN backup. I'll describe my scenario below, any doubt's feel free to ask me: - 2 Production nodes using RHEL 5.3 connected to SAN - 1 IBM DS4700 storage with 14 FC disks (13 raid5 + 1 hot-spare) - Each node have a emulex dual-port HBA - 2 Brocade TotalStorage fiber switches, with coherent zonings - Library is a IBM TS3100 LTO4 SAN, with only one drive Ok, i'm using TSM with LAN backups for a week, the transfer rate comes among 60~80GB per hour. The tape has a nominal write speed of 120MB/s that says me 430GB/h, ok.. this is a *nominal* speed, I was expecting something near 300GB/h, good enough for me. For these tests I've configured TSM Server on node1 and TSM Client on node2. Each node has 6 1Gbps NIC's, 4 to public network 2 for cluster-heartbeat + backups. With this setup I've got, as said above, 60~80GB/h. So, I decide to setup TSM for SAN.. Download a lot of docs/books and start reading/doing. I end up with: - Fabric zoning changes, isolating a HBA port on each dual-port cards for nodes - Installed StorageAgent on node2 - Installed TSM Client on node2 - - Configured the client to connect on storageagent - Redefine all my TSM Server setup - - Make Library as shared=yes, configure storage agent as a server, configure path, configure datawritepath=lanfree and so on... Ok, time to backup. And, for my surprise i got only ~500MB per minute.. I thought it very strange.. I've no idea what is the problem, I've tried with compression and without, DATAREADPATH=LAN and SAN, VirtualMountPoints or Physical ... Does anybody have an idea the reason of that bad transfer rate?? The servers are Samba Servers, with all kind of file and all size As I said before, I'm really new with TSM so is really possible that I've misconfigured something :) Anyway, thanks in advance, and any help will be appreciated. -- Flávio do Carmo Júnior aka waKKu Florianópolis/SC - Brazil No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.38/2274 - Release Date: 07/31/09 05:58:00
Database Cache Hit Ratio
Hello, In one of my TSM report, I see it says my database cash hit Ratio too high. What is the normal ratio? What do I have to do to bring it down or ways to resolve this? Thank you for your help. Issue Condition Recommendation The database utilization is too high. 85.4 85 Consider a database space trigger or increase the size of the database. [-] Custom Summary Item Results Administrative Schedules Successful: 6 Administrative Schedules Errors: 0 Administrative Schedules Failed: 0 Administrative Schedules Missed: 0 Client Schedules Completed with No Errors: 55 Client Schedules Completed with Skipped Files: 35 Client Schedules Completed with Warnings: 0 Client Schedules Completed with Errors: 0 Client Schedules Failed: 0 Client Schedules Missed: 0 Scheduled Nodes with empty contact information: 87 Total GB Backed Up: 1285.66 Total GB Restored: Total GB Archived: 160.81 Total GB Retrieved: 0.01 Last Database Backup: 2009-08-01 00:46 % Database Utilization: 85.4 Database Cache Hit Ratio: 99.1 % Maximum Recovery Log Utilization: 14.3 Disk Pool Total MB: 2007040.0 % Disk Pool Utilization: 16.1 Disk volumes offline: 0 Number of drives: 7 Number of tape mounts: 144 Total number of minutes tapes were mounted: 3855 Total number of volumes: 333 Number of scratch volumes: 57 Number of unavailable volumes: 0 Number of read-only volumes: 20 Volume Reclamation warning: 126 Avy Wong Business Continuity Administrator Mohegan Sun 1 Mohegan Sun Blvd Uncasville, CT 06382 (860)862-8164 (cell) (860)961-6976
Re: TSM 5.5 using LANFree backup really slow
What is the nature of the data that you're backing up here? Large files/objects or lots of smaller files/objects? The classic recommendation for SAN backups direct to tape is that it's best used only for larger objects (eg database backups, TDPs or disk image etc) to ensure optimal streaming to tape. Backing up a file server with millions of individual small files direct to tape won't always produce a great overall throughput figure. /David Mc London, UK Sent from my iPhone On 1 Aug 2009, at 15:37, Flavio Junior bil...@gmail.com wrote: Hi Steve, thanks for answering... Yeah, the destination is tape, i see on StorageAgent console the connection to server and opening tape, during the backup I see the bytes increasing on Backup by LANFree: (but, this value is ALWAYS lower than the value of normal backup... i don't know if this is normal). VALIDATE LANFREE node policyset shows me that the node is capable to LANFree backups. During the backup (using webclient interface) I can see two speeds, well.. A screenshot of backup report is at link: http://img18.imageshack.us/img18/9278/backupz.jpg The screenshot is in portuguese but I think is easy to you guess what is each value... Time: 5:45 Hours Size: 181 Gb No compression Rede == Network Agregada == Aggregate, combination... (not sure how TSM show it) Ok, this is what I have... If need more info just let me know. Thanks again. -- Flávio do Carmo Júnior aka waKKu Florianópolis/SC - Brazil On Sat, Aug 1, 2009 at 3:51 AM, Steven Harrisst...@stevenharris.info wrote: Flavio Is the destination of your backup pool tape? Have you confirmed with VALIDATE LANFREE on the server Are you seeing a proxied by message in the client log to indicate that the storage agent is in use? Is the *network* path configured as 1Gb - I'm suspecting you have a LAN backup and that the lan path has issues. HTH Regards Steve Steven Harris. TSM Admin, looking for work, Sydney Australia Flavio Junior wrote: Hi folks, I'm here, again, asking for help :) More I read, more doubts I have. I'm now implementing a TSM For SAN setup, to performance comparison with my currently LAN backup. I'll describe my scenario below, any doubt's feel free to ask me: - 2 Production nodes using RHEL 5.3 connected to SAN - 1 IBM DS4700 storage with 14 FC disks (13 raid5 + 1 hot-spare) - Each node have a emulex dual-port HBA - 2 Brocade TotalStorage fiber switches, with coherent zonings - Library is a IBM TS3100 LTO4 SAN, with only one drive Ok, i'm using TSM with LAN backups for a week, the transfer rate comes among 60~80GB per hour. The tape has a nominal write speed of 120MB/s that says me 430GB/h, ok.. this is a *nominal* speed, I was expecting something near 300GB/h, good enough for me. For these tests I've configured TSM Server on node1 and TSM Client on node2. Each node has 6 1Gbps NIC's, 4 to public network 2 for cluster-heartbeat + backups. With this setup I've got, as said above, 60~80GB/h. So, I decide to setup TSM for SAN.. Download a lot of docs/books and start reading/doing. I end up with: - Fabric zoning changes, isolating a HBA port on each dual-port cards for nodes - Installed StorageAgent on node2 - Installed TSM Client on node2 - - Configured the client to connect on storageagent - Redefine all my TSM Server setup - - Make Library as shared=yes, configure storage agent as a server, configure path, configure datawritepath=lanfree and so on... Ok, time to backup. And, for my surprise i got only ~500MB per minute.. I thought it very strange.. I've no idea what is the problem, I've tried with compression and without, DATAREADPATH=LAN and SAN, VirtualMountPoints or Physical ... Does anybody have an idea the reason of that bad transfer rate?? The servers are Samba Servers, with all kind of file and all size As I said before, I'm really new with TSM so is really possible that I've misconfigured something :) Anyway, thanks in advance, and any help will be appreciated. -- Flávio do Carmo Júnior aka waKKu Florianópolis/SC - Brazil --- --- -- No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.38/2274 - Release Date: 07/31/09 05:58:00
Re: Database Cache Hit Ratio
The Recommendation you cite is for database utilization, not for cache hit ratio. Your cache hit ratio looks healthy, at 99.1%. I believe that the warning for database utilization is a warning that your database is getting full, and that you need to do something before it fills up completely. ..Paul At 10:58 AM 8/1/2009, Avy Wong wrote: Hello, In one of my TSM report, I see it says my database cash hit Ratio too high. What is the normal ratio? What do I have to do to bring it down or ways to resolve this? Thank you for your help. Issue Condition Recommendation The database utilization is too high. 85.4 85 Consider a database space trigger or increase the size of the database. [-] Custom Summary Item Results Administrative Schedules Successful: 6 Administrative Schedules Errors: 0 Administrative Schedules Failed: 0 Administrative Schedules Missed: 0 Client Schedules Completed with No Errors: 55 Client Schedules Completed with Skipped Files: 35 Client Schedules Completed with Warnings: 0 Client Schedules Completed with Errors: 0 Client Schedules Failed: 0 Client Schedules Missed: 0 Scheduled Nodes with empty contact information: 87 Total GB Backed Up: 1285.66 Total GB Restored: Total GB Archived: 160.81 Total GB Retrieved: 0.01 Last Database Backup: 2009-08-01 00:46 % Database Utilization: 85.4 Database Cache Hit Ratio: 99.1 % Maximum Recovery Log Utilization: 14.3 Disk Pool Total MB: 2007040.0 % Disk Pool Utilization: 16.1 Disk volumes offline: 0 Number of drives: 7 Number of tape mounts: 144 Total number of minutes tapes were mounted: 3855 Total number of volumes: 333 Number of scratch volumes: 57 Number of unavailable volumes: 0 Number of read-only volumes: 20 Volume Reclamation warning: 126 Avy Wong Business Continuity Administrator Mohegan Sun 1 Mohegan Sun Blvd Uncasville, CT 06382 (860)862-8164 (cell) (860)961-6976 -- Paul ZarnowskiPh: 607-255-4757 Manager, Storage Services Fx: 607-255-8521 719 Rhodes Hall, Ithaca, NY 14853-3801Em: p...@cornell.edu
Re: Database Cache Hit Ratio
Any recommendations to resolve this ? Avy Wong Business Continuity Administrator Mohegan Sun 1 Mohegan Sun Blvd Uncasville, CT 06382 (860)862-8164 (cell) (860)961-6976 Paul Zarnowski p...@cornell.edu To Sent by: ADSM: ADSM-L@VM.MARIST.EDU Dist Stor cc Manager ads...@vm.marist Subject .EDU Re: [ADSM-L] Database Cache Hit Ratio 08/01/2009 11:08 AM Please respond to ADSM: Dist Stor Manager ads...@vm.marist .EDU The Recommendation you cite is for database utilization, not for cache hit ratio. Your cache hit ratio looks healthy, at 99.1%. I believe that the warning for database utilization is a warning that your database is getting full, and that you need to do something before it fills up completely. ..Paul At 10:58 AM 8/1/2009, Avy Wong wrote: Hello, In one of my TSM report, I see it says my database cash hit Ratio too high. What is the normal ratio? What do I have to do to bring it down or ways to resolve this? Thank you for your help. Issue Condition Recommendation The database utilization is too high. 85.4 85 Consider a database space trigger or increase the size of the database. [-] Custom Summary Item Results Administrative Schedules Successful: 6 Administrative Schedules Errors: 0 Administrative Schedules Failed: 0 Administrative Schedules Missed: 0 Client Schedules Completed with No Errors: 55 Client Schedules Completed with Skipped Files: 35 Client Schedules Completed with Warnings: 0 Client Schedules Completed with Errors: 0 Client Schedules Failed: 0 Client Schedules Missed: 0 Scheduled Nodes with empty contact information: 87 Total GB Backed Up: 1285.66 Total GB Restored: Total GB Archived: 160.81 Total GB Retrieved: 0.01 Last Database Backup: 2009-08-01 00:46 % Database Utilization: 85.4 Database Cache Hit Ratio: 99.1 % Maximum Recovery Log Utilization: 14.3 Disk Pool Total MB: 2007040.0 % Disk Pool Utilization: 16.1 Disk volumes offline: 0 Number of drives: 7 Number of tape mounts: 144 Total number of minutes tapes were mounted: 3855 Total number of volumes: 333 Number of scratch volumes: 57 Number of unavailable volumes: 0 Number of read-only volumes: 20 Volume Reclamation warning: 126 Avy Wong Business Continuity Administrator Mohegan Sun 1 Mohegan Sun Blvd Uncasville, CT 06382 (860)862-8164 (cell) (860)961-6976 -- Paul ZarnowskiPh: 607-255-4757 Manager, Storage Services Fx: 607-255-8521 719 Rhodes Hall, Ithaca, NY 14853-3801Em: p...@cornell.edu
Re: Database Cache Hit Ratio
Increase the capacity of the database. At 11:13 AM 8/1/2009, Avy Wong wrote: Any recommendations to resolve this ? Avy Wong Business Continuity Administrator Mohegan Sun 1 Mohegan Sun Blvd Uncasville, CT 06382 (860)862-8164 (cell) (860)961-6976 Paul Zarnowski p...@cornell.edu To Sent by: ADSM: ADSM-L@VM.MARIST.EDU Dist Stor cc Manager ads...@vm.marist Subject .EDU Re: [ADSM-L] Database Cache Hit Ratio 08/01/2009 11:08 AM Please respond to ADSM: Dist Stor Manager ads...@vm.marist .EDU The Recommendation you cite is for database utilization, not for cache hit ratio. Your cache hit ratio looks healthy, at 99.1%. I believe that the warning for database utilization is a warning that your database is getting full, and that you need to do something before it fills up completely. ..Paul At 10:58 AM 8/1/2009, Avy Wong wrote: Hello, In one of my TSM report, I see it says my database cash hit Ratio too high. What is the normal ratio? What do I have to do to bring it down or ways to resolve this? Thank you for your help. Issue Condition Recommendation The database utilization is too high. 85.4 85 Consider a database space trigger or increase the size of the database. [-] Custom Summary Item Results Administrative Schedules Successful: 6 Administrative Schedules Errors: 0 Administrative Schedules Failed: 0 Administrative Schedules Missed: 0 Client Schedules Completed with No Errors: 55 Client Schedules Completed with Skipped Files: 35 Client Schedules Completed with Warnings: 0 Client Schedules Completed with Errors: 0 Client Schedules Failed: 0 Client Schedules Missed: 0 Scheduled Nodes with empty contact information: 87 Total GB Backed Up: 1285.66 Total GB Restored: Total GB Archived: 160.81 Total GB Retrieved: 0.01 Last Database Backup: 2009-08-01 00:46 % Database Utilization: 85.4 Database Cache Hit Ratio: 99.1 % Maximum Recovery Log Utilization: 14.3 Disk Pool Total MB: 2007040.0 % Disk Pool Utilization: 16.1 Disk volumes offline: 0 Number of drives: 7 Number of tape mounts: 144 Total number of minutes tapes were mounted: 3855 Total number of volumes: 333 Number of scratch volumes: 57 Number of unavailable volumes: 0 Number of read-only volumes: 20 Volume Reclamation warning: 126 Avy Wong Business Continuity Administrator Mohegan Sun 1 Mohegan Sun Blvd Uncasville, CT 06382 (860)862-8164 (cell) (860)961-6976 -- Paul ZarnowskiPh: 607-255-4757 Manager, Storage Services Fx: 607-255-8521 719 Rhodes Hall, Ithaca, NY 14853-3801Em: p...@cornell.edu -- Paul ZarnowskiPh: 607-255-4757 Manager, Storage Services Fx: 607-255-8521 719 Rhodes Hall, Ithaca, NY 14853-3801Em: p...@cornell.edu
Re: Database Cache Hit Ratio
On Aug 1, 2009, at 11:13 AM, Avy Wong wrote: Any recommendations to resolve this ? In ADSM QuickFacts I provide a list of known contributors to database utilization. Do obvious things such as assure that expirations are running to completion, that obsolete filespaces are not hanging around, etc. Rigorous ongoing administration prevents most problems. Richard Sims
Re: TSM 5.5 using LANFree backup really slow
Hi David, Well FileServer is exact what i'm backup ;/ A lot of files with sizes between 100k and 20Mb. So, there is nothing that can be done to improve this, at least for 100GB/h? Just to point my head toward right direction.. LANFree backups or LAN backups. Thanks, -- Flávio do Carmo Júnior aka waKKu Florianópolis/SC - Brazil On Sat, Aug 1, 2009 at 12:02 PM, David McClellanddavid.mcclell...@networkc.co.uk wrote: What is the nature of the data that you're backing up here? Large files/objects or lots of smaller files/objects? The classic recommendation for SAN backups direct to tape is that it's best used only for larger objects (eg database backups, TDPs or disk image etc) to ensure optimal streaming to tape. Backing up a file server with millions of individual small files direct to tape won't always produce a great overall throughput figure. /David Mc London, UK Sent from my iPhone On 1 Aug 2009, at 15:37, Flavio Junior bil...@gmail.com wrote: Hi Steve, thanks for answering... Yeah, the destination is tape, i see on StorageAgent console the connection to server and opening tape, during the backup I see the bytes increasing on Backup by LANFree: (but, this value is ALWAYS lower than the value of normal backup... i don't know if this is normal). VALIDATE LANFREE node policyset shows me that the node is capable to LANFree backups. During the backup (using webclient interface) I can see two speeds, well.. A screenshot of backup report is at link: http://img18.imageshack.us/img18/9278/backupz.jpg The screenshot is in portuguese but I think is easy to you guess what is each value... Time: 5:45 Hours Size: 181 Gb No compression Rede == Network Agregada == Aggregate, combination... (not sure how TSM show it) Ok, this is what I have... If need more info just let me know. Thanks again. -- Flávio do Carmo Júnior aka waKKu Florianópolis/SC - Brazil On Sat, Aug 1, 2009 at 3:51 AM, Steven Harrisst...@stevenharris.info wrote: Flavio Is the destination of your backup pool tape? Have you confirmed with VALIDATE LANFREE on the server Are you seeing a proxied by message in the client log to indicate that the storage agent is in use? Is the *network* path configured as 1Gb - I'm suspecting you have a LAN backup and that the lan path has issues. HTH Regards Steve Steven Harris. TSM Admin, looking for work, Sydney Australia Flavio Junior wrote: Hi folks, I'm here, again, asking for help :) More I read, more doubts I have. I'm now implementing a TSM For SAN setup, to performance comparison with my currently LAN backup. I'll describe my scenario below, any doubt's feel free to ask me: - 2 Production nodes using RHEL 5.3 connected to SAN - 1 IBM DS4700 storage with 14 FC disks (13 raid5 + 1 hot-spare) - Each node have a emulex dual-port HBA - 2 Brocade TotalStorage fiber switches, with coherent zonings - Library is a IBM TS3100 LTO4 SAN, with only one drive Ok, i'm using TSM with LAN backups for a week, the transfer rate comes among 60~80GB per hour. The tape has a nominal write speed of 120MB/s that says me 430GB/h, ok.. this is a *nominal* speed, I was expecting something near 300GB/h, good enough for me. For these tests I've configured TSM Server on node1 and TSM Client on node2. Each node has 6 1Gbps NIC's, 4 to public network 2 for cluster-heartbeat + backups. With this setup I've got, as said above, 60~80GB/h. So, I decide to setup TSM for SAN.. Download a lot of docs/books and start reading/doing. I end up with: - Fabric zoning changes, isolating a HBA port on each dual-port cards for nodes - Installed StorageAgent on node2 - Installed TSM Client on node2 - - Configured the client to connect on storageagent - Redefine all my TSM Server setup - - Make Library as shared=yes, configure storage agent as a server, configure path, configure datawritepath=lanfree and so on... Ok, time to backup. And, for my surprise i got only ~500MB per minute.. I thought it very strange.. I've no idea what is the problem, I've tried with compression and without, DATAREADPATH=LAN and SAN, VirtualMountPoints or Physical ... Does anybody have an idea the reason of that bad transfer rate?? The servers are Samba Servers, with all kind of file and all size As I said before, I'm really new with TSM so is really possible that I've misconfigured something :) Anyway, thanks in advance, and any help will be appreciated. -- Flávio do Carmo Júnior aka waKKu Florianópolis/SC - Brazil No virus found in this incoming message. Checked by AVG - www.avg.com Version: 8.5.392 / Virus Database: 270.13.38/2274 - Release Date: 07/31/09 05:58:00
Re: TSM 5.5 using LANFree backup really slow
LAN Free will be poor with this kind of data. What is your actual backup restore requirement? What's your backup window, how much data and how much changed data per day. Steven Steven Langdale Global Information Services EAME SAN/Storage Planning and Implementation ( Phone : +44 (0)1733 584175 ( Mob: +44 (0)7876 216782 ü Conference: +44 (0)208 609 7400 Code: 331817 + Email: steven.langd...@cat.com Flavio Junior bil...@gmail.com Sent by: ADSM: Dist Stor Manager ADSM-L@VM.MARIST.EDU 01/08/2009 17:10 Please respond to ADSM: Dist Stor Manager ADSM-L@VM.MARIST.EDU To ADSM-L@VM.MARIST.EDU cc Subject Re: [ADSM-L] TSM 5.5 using LANFree backup really slow Caterpillar: Confidential Green Retain Until: 31/08/2009 Hi David, Well FileServer is exact what i'm backup ;/ A lot of files with sizes between 100k and 20Mb. So, there is nothing that can be done to improve this, at least for 100GB/h? Just to point my head toward right direction.. LANFree backups or LAN backups. Thanks, -- Flávio do Carmo Júnior aka waKKu Florianópolis/SC - Brazil On Sat, Aug 1, 2009 at 12:02 PM, David McClellanddavid.mcclell...@networkc.co.uk wrote: What is the nature of the data that you're backing up here? Large files/objects or lots of smaller files/objects? The classic recommendation for SAN backups direct to tape is that it's best used only for larger objects (eg database backups, TDPs or disk image etc) to ensure optimal streaming to tape. Backing up a file server with millions of individual small files direct to tape won't always produce a great overall throughput figure. /David Mc London, UK Sent from my iPhone On 1 Aug 2009, at 15:37, Flavio Junior bil...@gmail.com wrote: Hi Steve, thanks for answering... Yeah, the destination is tape, i see on StorageAgent console the connection to server and opening tape, during the backup I see the bytes increasing on Backup by LANFree: (but, this value is ALWAYS lower than the value of normal backup... i don't know if this is normal). VALIDATE LANFREE node policyset shows me that the node is capable to LANFree backups. During the backup (using webclient interface) I can see two speeds, well.. A screenshot of backup report is at link: http://img18.imageshack.us/img18/9278/backupz.jpg The screenshot is in portuguese but I think is easy to you guess what is each value... Time: 5:45 Hours Size: 181 Gb No compression Rede == Network Agregada == Aggregate, combination... (not sure how TSM show it) Ok, this is what I have... If need more info just let me know. Thanks again. -- Flávio do Carmo Júnior aka waKKu Florianópolis/SC - Brazil On Sat, Aug 1, 2009 at 3:51 AM, Steven Harrisst...@stevenharris.info wrote: Flavio Is the destination of your backup pool tape? Have you confirmed with VALIDATE LANFREE on the server Are you seeing a proxied by message in the client log to indicate that the storage agent is in use? Is the *network* path configured as 1Gb - I'm suspecting you have a LAN backup and that the lan path has issues. HTH Regards Steve Steven Harris. TSM Admin, looking for work, Sydney Australia Flavio Junior wrote: Hi folks, I'm here, again, asking for help :) More I read, more doubts I have. I'm now implementing a TSM For SAN setup, to performance comparison with my currently LAN backup. I'll describe my scenario below, any doubt's feel free to ask me: - 2 Production nodes using RHEL 5.3 connected to SAN - 1 IBM DS4700 storage with 14 FC disks (13 raid5 + 1 hot-spare) - Each node have a emulex dual-port HBA - 2 Brocade TotalStorage fiber switches, with coherent zonings - Library is a IBM TS3100 LTO4 SAN, with only one drive Ok, i'm using TSM with LAN backups for a week, the transfer rate comes among 60~80GB per hour. The tape has a nominal write speed of 120MB/s that says me 430GB/h, ok.. this is a *nominal* speed, I was expecting something near 300GB/h, good enough for me. For these tests I've configured TSM Server on node1 and TSM Client on node2. Each node has 6 1Gbps NIC's, 4 to public network 2 for cluster-heartbeat + backups. With this setup I've got, as said above, 60~80GB/h. So, I decide to setup TSM for SAN.. Download a lot of docs/books and start reading/doing. I end up with: - Fabric zoning changes, isolating a HBA port on each dual-port cards for nodes - Installed StorageAgent on node2 - Installed TSM Client on node2 - - Configured the client to connect on storageagent - Redefine all my TSM Server setup - - Make Library as shared=yes, configure storage agent as a server, configure path, configure datawritepath=lanfree and so on... Ok, time to backup. And, for my surprise i got only ~500MB per minute.. I thought it very strange.. I've no idea what is the problem, I've tried with compression and without, DATAREADPATH=LAN and SAN, VirtualMountPoints or Physical ... Does anybody have an idea the reason of that bad transfer rate?? The
Watch-for: TS3310 firmware upgrade when used with Qlogic HBAs on x86
I just had two installations with upgrades of TS3310 firmware to level 570 (the most current). They had QLE2460/2 and QLA2460/2 HBAs. One was TSM for Windows 5.5.1, the other TSM for Linux 6.1 on x86. In both cases the installations were getting IO errors and could not get the library and/or drives to work (ANR8311E errno=5 or just IO error accessing...). You will find that QLE/A 2460/2 cards are not supported with firmware 570 if you use the HBA search tool on the IBM support site: http://www-03.ibm.com/systems/support/storage/config/hba/index.wss If you downlevel the TS3310 library firmware to 520 and the corresponding drive firmware for LTOx, the library/drives will work fine. You will need to reboot the server after the firmware change and possibly delete the library and redefine the library plus check in all scratches (search=yes) in the library followed by the check in of the private tapes (search=yes). Joerg Pohlmann 250-245-9863
