long restores

I'm restoring level-0 backups from one server to another (IDS 7.24 & HPUX 11 on both). The archive only takes a couple hours, but the restores take over 6 to complete. When finished, the database comes online fine with no problems. I just can't figure out why the restore takes so long. Any ideas?

Need more info dude.... Level-0 backup is from what tool? (ontape, onarchive, onbar?) i.e. if onbar backup, was it mult-threaded (bar_max(?) in onconfig in combo with backup software)? i.e. we use veritas netbackup and run 12 threads, but on a restore (alternate client to another server such as you are doing), we would use less than 12 threads to avoid tape contension. Is your network dropping into half-duplex mode?.... Please provide more info to the mail group... Thanks, Norma Jean -----Original Message----- From: forum.subscriber@iiug.org [mailto:forum.subscriber@iiug.org] On Behalf Of DAN DEHAAN Sent: Tuesday, December 28, 2004 3:30 PM To: ids@iiug.org Subject: long restores [3926] I'm restoring level-0 backups from one server to another (IDS 7.24 & HPUX 11 on both). The archive only takes a couple hours, but the restores take over 6 to complete. When finished, the database comes online fine with no problems. I just can't figure out why the restore takes so long. Any ideas? ============================================================ The information contained in this message may be privileged and confidential and protected from disclosure. If the reader of this message is not the intended recipient, or an employee or agent responsible for delivering this message to the intended recipient, you are hereby notified that any reproduction, dissemination or distribution of this communication is strictly prohibited. If you have received this communication in error, please notify us immediately by replying to the message and deleting it from your computer. Thank you. Tellabs ============================================================

Dan Look at disks as well, are they attached n the same way? are they the same size/speed? are they both raw/cooked spaces? Have you monitored the restore to see if one spaces is causing more problems than the others? Keith -> -----Original Message----- -> From: Sebastian, .... [mailto:NormaJean.Sebastian@tellabs.com] -> Sent: Wednesday, December 29, 2004 2:37 AM -> To: ids@iiug.org -> Subject: RE: long restores [3927] -> -> -> -> Need more info dude.... -> -> Level-0 backup is from what tool? (ontape, onarchive, onbar?) -> -> i.e. if onbar backup, was it mult-threaded (bar_max(?) in onconfig in -> combo with backup software)? i.e. we use veritas netbackup -> and run 12 -> threads, but on a restore (alternate client to another server such as -> you are doing), we would use less than 12 threads to avoid tape -> contension. -> -> Is your network dropping into half-duplex mode?.... -> -> Please provide more info to the mail group... -> Thanks, -> Norma Jean -> -> -> -> -----Original Message----- -> From: forum.subscriber@iiug.org [mailto:forum.subscriber@iiug.org] On -> Behalf Of DAN DEHAAN -> Sent: Tuesday, December 28, 2004 3:30 PM -> To: ids@iiug.org -> Subject: long restores [3926] -> -> I'm restoring level-0 backups from one server to another (IDS 7.24 & -> HPUX 11 on both). The archive only takes a couple hours, but the -> restores take over 6 to complete. When finished, the database comes -> online fine with no problems. I just can't figure out why -> the restore -> takes so long. Any ideas? -> ============================================================ -> The information contained in this message may be privileged -> and confidential and protected from disclosure. If the reader -> of this message is not the intended recipient, or an employee -> or agent responsible for delivering this message to the -> intended recipient, you are hereby notified that any reproduction, -> dissemination or distribution of this communication is strictly -> prohibited. If you have received this communication in error, -> please notify us immediately by replying to the message and -> deleting it from your computer. Thank you. Tellabs -> ============================================================ -> -> ******************************************************************************** ** This message is sent in strict confidence for the addressee only. It may contain legally privileged information. The contents are not to be disclosed to anyone other than the addressee. Unauthorised recipients are requested to preserve this confidentiality and to advise the sender immediately of any error in transmission. This footnote also confirms that this email message has been swept for the presence of computer viruses, however we cannot guarantee that this message is free from such problems. ******************************************************************************** **

Hi I take it you are using ontape with remote tape option. Before we migrate to onbar we had the same problem. So for backup I use remote tape but for restore I moved the tape so it will be local. Uri -------------------------------------------------------- I'm restoring level-0 backups from one server to another (IDS 7.24 & HPUX 11 on both). The archive only takes a couple hours, but the restores take over 6 to complete. When finished, the database comes online fine with no problems. I just can't figure out why the restore takes so long. Any ideas?

I guess I was a little vague, wasn't I. I'm using ontape from a local tape drive. Another reply got me looking at my logical volumes and I noticed I had linked to the cooked volumes and not raw. I'm trying another restore. -----Original Message----- From: Sebastian, Norma J. [mailto:NormaJean.Sebastian@tellabs.com] Sent: Tuesday, December 28, 2004 7:32 PM To: Dan Dehaan; ids@iiug.org Subject: RE: long restores [3926] Need more info dude.... Level-0 backup is from what tool? (ontape, onarchive, onbar?) i.e. if onbar backup, was it mult-threaded (bar_max(?) in onconfig in combo with backup software)? i.e. we use veritas netbackup and run 12 threads, but on a restore (alternate client to another server such as you are doing), we would use less than 12 threads to avoid tape contension. Is your network dropping into half-duplex mode?.... Please provide more info to the mail group... Thanks, Norma Jean -----Original Message----- From: forum.subscriber@iiug.org [mailto:forum.subscriber@iiug.org] On Behalf Of DAN DEHAAN Sent: Tuesday, December 28, 2004 3:30 PM To: ids@iiug.org Subject: long restores [3926] I'm restoring level-0 backups from one server to another (IDS 7.24 & HPUX 11 on both). The archive only takes a couple hours, but the restores take over 6 to complete. When finished, the database comes online fine with no problems. I just can't figure out why the restore takes so long. Any ideas? ============================================================ The information contained in this message may be privileged and confidential and protected from disclosure. If the reader of this message is not the intended recipient, or an employee or agent responsible for delivering this message to the intended recipient, you are hereby notified that any reproduction, dissemination or distribution of this communication is strictly prohibited. If you have received this communication in error, please notify us immediately by replying to the message and deleting it from your computer. Thank you. Tellabs ============================================================

In general, we always budget twice the time to restore a file as it does to back it up. That is just how long it takes. You can always speed up the general process by making sure all the restore tapes are available (with a tape silo) and that no other processes are taking up the network. -----Original Message----- From: forum.subscriber@iiug.org [mailto:forum.subscriber@iiug.org] On Behalf Of DAN DEHAAN Sent: Tuesday, December 28, 2004 4:30 PM To: ids@iiug.org Subject: long restores [3926] I'm restoring level-0 backups from one server to another (IDS 7.24 & HPUX 11 on both). The archive only takes a couple hours, but the restores take over 6 to complete. When finished, the database comes online fine with no problems. I just can't figure out why the restore takes so long. Any ideas?

Hi, there are several factors that can make the restore (considerably) longer than the backup. The 2 most likely are: - when using cooked files (for chunks) they must exist before starting the restore. But if they only exist and the restore needs to "blow them up" to respective size, then this can take considerable time (the OS needs to create the necessary inodes, etc.). I hope you're using raw devices. Otherwise you can "blow up" the cooked chunk files before starting the restore, e.g. using a command like "dd if=/dev/zero of=<chunk file> bs=2048 count=<size in pages>" to pre-allocate the disk space for cooked chunk files. - after physical restore (copying backup data to disk), a logical restore needs to be done. This can be actual restore of logical log files from tape and rollforward of the log records therein. Even if no log files are restored, the server still needs to check for open transactions (i.e. that were open at backup time) and roll them back to make the restore logically consistent. This rolling back of open transactions can take quite a while. You may be able to figure out from the IDS message log (online.log), which phase of the restore takes what time to complete. That way you can figure out, where the time is used and where something can be improved ... Regards, Martin -- Martin Fuerderer IBM Informix Development Munich, Germany Data Management Solutions forum.subscriber@iiug.org wrote on 03.01.2005 15:35:50: > In general, we always budget twice the time to restore a file as it does > to back it up. That is just how long it takes. You can always speed up > the general process by making sure all the restore tapes are available > (with a tape silo) and that no other processes are taking up the > network. > > > -----Original Message----- > From: forum.subscriber@iiug.org [mailto:forum.subscriber@iiug.org] On > Behalf Of DAN DEHAAN > Sent: Tuesday, December 28, 2004 4:30 PM > To: ids@iiug.org > Subject: long restores [3926] > > > I'm restoring level-0 backups from one server to another (IDS 7.24 & > HPUX 11 on both). The archive only takes a couple hours, but the > restores take over 6 to complete. When finished, the database comes > online fine with no problems. I just can't figure out why the restore > takes so long. Any ideas? > > >

Martin, Based on your statements, is it safe to assume that there is not much to do in the onconfig file to help reduce the length of time for a restore? I have read thru the docs and can't find anything to change in the onconfig. NUMAIOVPS is set properly. I too have a long restore problem, but our 1 hour backup takes 18-20 hours to restore. We are using cooked files (setup long before I got here), but the number of chunks we have is rather large for our system. We have about 55GB of data on 173 chunks. The performance docs say you should have 1 chunk per disk. From what I hear, 9.40 allows you to 'condense' chunks during a restore. I hope that is true. Thanks.