This is a discussion on question about OnDemand's abilities? within the AIX Operating System forums, part of the Unix Operating Systems category; --> A question came up yesterday about the possible abilities of OnDemand. Currently we have OnDemand in house storing various ...
| |||||||
| Register | FAQ | Members List | Calendar | Search | Today's Posts | Mark Forums Read |
| ||||
| A question came up yesterday about the possible abilities of OnDemand. Currently we have OnDemand in house storing various documents for legal (I think) reasons. We are talking about having a HSM for older user documents. The users have windows workstations with windows file servers. The OnDemand system is on AIX. Can OnDemand act as a HSM for older documents, and if so how do I merge it with the windows environment? Mike |
| |||
| Mike <mikee@mikee.ath.cx> wrote: > A question came up yesterday about the possible abilities > of OnDemand. Currently we have OnDemand in house storing > various documents for legal (I think) reasons. We are talking > about having a HSM for older user documents. The users have > windows workstations with windows file servers. The OnDemand > system is on AIX. Can OnDemand act as a HSM for older documents, > and if so how do I merge it with the windows environment? OnDemand may have changed since I last worked with it (haven't they relabeled it to DB2 ContentManager OnDemand or something?). OnDemand uses TSM (formerly ADSM) for storage management, and you can specify migration policies to move documents to cheaper mass storage. Maybe this is not hierarchical storage management as you expect it, since documents are not migrated back to fast storage when you access them again. Also, as an archive system, it will not allow you to modify archived files. About merging with the windows environment, I don't know, but they seem to have client software now that allows viewing the document in a web browser. My personal experience is that OnDemand works best with large spool files, for which it was originally written. Loading lots of small files may prove to be painful. OnDemand is good software IMO, but as always software works best if you use it as intended. Maybe the documentation at http://www-306.ibm.com/software/data...p/library.html can enlighten you further. One last hint: If IBM tries to sell you DB2 ContentManager, that is a different product, and although this product is designed to handle small files, DO NOT BUY IT. Just my 2 cents. Yours, Laurenz Albe |
| |||
| In article <ciepsi$5af$1@at-vie-newsmaster01.nextra.at>, Laurenz Albe wrote: > Mike <mikee@mikee.ath.cx> wrote: >> A question came up yesterday about the possible abilities >> of OnDemand. Currently we have OnDemand in house storing >> various documents for legal (I think) reasons. We are talking >> about having a HSM for older user documents. The users have >> windows workstations with windows file servers. The OnDemand >> system is on AIX. Can OnDemand act as a HSM for older documents, >> and if so how do I merge it with the windows environment? > > OnDemand may have changed since I last worked with it (haven't they > relabeled it to DB2 ContentManager OnDemand or something?). > > OnDemand uses TSM (formerly ADSM) for storage management, and you > can specify migration policies to move documents to cheaper mass > storage. Maybe this is not hierarchical storage management as you > expect it, since documents are not migrated back to fast storage > when you access them again. Also, as an archive system, it will not > allow you to modify archived files. > > About merging with the windows environment, I don't know, but they > seem to have client software now that allows viewing the document in > a web browser. > > My personal experience is that OnDemand works best with large spool > files, for which it was originally written. Loading lots of small > files may prove to be painful. > > OnDemand is good software IMO, but as always software works best if > you use it as intended. > > Maybe the documentation at > http://www-306.ibm.com/software/data...p/library.html > can enlighten you further. > > One last hint: If IBM tries to sell you DB2 ContentManager, that is > a different product, and although this product is designed to handle > small files, DO NOT BUY IT. Just my 2 cents. > > Yours, > Laurenz Albe Thanks for the link! |
| |||
| Mike <mikee@mikee.ath.cx> wrote: >>> Currently we have OnDemand in house storing >>> various documents for legal (I think) reasons. We are talking >>> about having a HSM for older user documents. One more thought: Since you have OnDemand, you'll probably have TSM too. TSM has hierarchical storage management capacities, maybe that's what you're looking for! Laurenz Albe |
| |||
| In article <cim4b8$7ac$1@at-vie-newsmaster01.nextra.at>, Laurenz Albe wrote: > Mike <mikee@mikee.ath.cx> wrote: >>>> Currently we have OnDemand in house storing >>>> various documents for legal (I think) reasons. We are talking >>>> about having a HSM for older user documents. > > One more thought: Since you have OnDemand, you'll probably have TSM too. > TSM has hierarchical storage management capacities, maybe that's what > you're looking for! > > Laurenz Albe I do and did think about TSM, but though it does do HSM, it doesn't do the management of staging older files out to slower storage. It seeems to work great for backups, but not for users. Mike |
| |||
| Mike <mikee@mikee.ath.cx> wrote: >> One more thought: Since you have OnDemand, you'll probably have TSM too. >> TSM has hierarchical storage management capacities, maybe that's what >> you're looking for! > > I do and did think about TSM, but though it does do HSM, it doesn't > do the management of staging older files out to slower storage. It seeems > to work great for backups, but not for users. This is getting off-topic, and there may be a misunderstanding, but what about defining a management class with def mgm <domain> <polset> hsm spacemgtech=auto automigno=30 migdest=slow Then it should migrate all files older than 30 days to a slow storage pool. Yours, Laurenz Albe |
| |||
| In article <cimmfg$l16$1@at-vie-newsmaster01.nextra.at>, Laurenz Albe wrote: > Mike <mikee@mikee.ath.cx> wrote: >>> One more thought: Since you have OnDemand, you'll probably have TSM too. >>> TSM has hierarchical storage management capacities, maybe that's what >>> you're looking for! >> >> I do and did think about TSM, but though it does do HSM, it doesn't >> do the management of staging older files out to slower storage. It seeems >> to work great for backups, but not for users. > > This is getting off-topic, and there may be a misunderstanding, but what > about defining a management class with > > def mgm <domain> <polset> hsm spacemgtech=auto automigno=30 migdest=slow > > Then it should migrate all files older than 30 days to a slow storage pool. > > Yours, > Laurenz Albe That's a great idea. How would the users get it back when they asked for it? Mike |
| ||||
| Mike <mikee@mikee.ath.cx> wrote: >> def mgm <domain> <polset> hsm spacemgtech=auto automigno=30 migdest=slow >> >> Then it should migrate all files older than 30 days to a slow storage pool. > > That's a great idea. How would the users get it back when they asked for it? Transparently by accessing the stub file on the client. It's some time since I've worked with TSM, so forgive if I am wrong, but AFAIK accessing a file from a TSM client works the same, even after the file has been moved to another, slower storage pool. So, after 30 days, your file is in storage pool 'slow', and there is a stub file on the client. If there is a nextstgpool 'slower' for this storage pool, then after some time the file ends up in 'slower'. The stub file on the client remains. When you access it, the file is fetched from 'slower'. Yours, Laurenz Albe |