... It's not that one, since I don't have a 64-bit system. However, after sending the email I realized the probable cause of the problem: I started using...
4570
Geert Uytterhoeven
geert@...
Sep 1, 2005 2:15 pm
... After some suggestions from Paul Bijnens, I extracted the following relevant ... So the estimate from the last level 0 was -1K. And the estimate from the...
4571
Geert Uytterhoeven
geert@...
Sep 4, 2005 11:57 am
... The following patch fixes the issue (at least in the planner output, a real dump is scheduled for tonight): 1. When using server-side estimates, failed...
4572
Joe Rhett
jrhett@...
Sep 7, 2005 8:26 am
This was discussed on amanda-users, but nobody has stepped up to address this. It's clearly a fairly major bug in the planner code... ... Here's the relevant...
4573
Joe Rhett
jrhett@...
Sep 7, 2005 8:28 am
I sent this to -users, and got nothing more than a "yeah". What will it take to get this fixed? ... Date: Fri, 26 Aug 2005 12:23:12 -0700 From: Joe Rhett...
4574
Paul Bijnens
paul.bijnens@...
Sep 7, 2005 9:27 am
... A patch attached. yeah :-) Patch is made against amanda version 2.4.5. This patch looses the nice column for the hostname part in the "FAILURE AND...
4575
Paul Bijnens
paul.bijnens@...
Sep 7, 2005 10:25 am
... Patch attached. yeah :-) The patch is against amanda 2.4.5. This solves the problem that a delayed-full-because-too-large turns into an...
4576
Gregor Ibic
gregor.ibic@...
Sep 7, 2005 11:03 am
I would like to know if this fixes or is related to my behaviour. I have LTO 200GB(400 compressed) and I try to backup 360 GB of data, and it goes to tape...
4577
Stefan G. Weichinger
sgw@...
Sep 7, 2005 11:48 am
At last I found some time again and have added recently posted patches to http://www.amanda.org/patches/ These patches have been posted on the...
4578
Joe Rhett
jrhett@...
Sep 7, 2005 4:15 pm
... And the patch will be provided when...? -- Joe Rhett senior geek meer.net...
4579
Paul Bijnens
paul.bijnens@...
Sep 7, 2005 5:14 pm
... The patch attached solves the last problem too. It patches amflush and planner (for autoflush) to skip all holdingdisk files that are larger than the size...
4580
Jean-Louis Martineau
martinea@...
Sep 7, 2005 5:25 pm
I Orion, I commited your tapeleft and updated amoverview patches. I will send another email for he taperwait patch. Stefan, could you update the xml...
4581
Jean-Louis Martineau
martinea@...
Sep 7, 2005 5:33 pm
I Orion, Your taperwait patch might be useful but it doesn't solve some of the problem in the taper scheduling. eg. 1- I don't want to use a tape every day,...
4582
Jean-Louis Martineau
martinea@...
Sep 7, 2005 6:01 pm
Hi Paul, ... This can happen anytime for any level if the estimate size is smaller that the dumped size. I think it should be fixed in the driver, not in...
4583
Jon LaBadie
jon@...
Sep 7, 2005 6:10 pm
... Paul, just a minor question that your wording prompts. Do you mean a "dump of a DLE" that is larger or the individual holding disk files. What I'm...
4584
Paul Bijnens
paul.bijnens@...
Sep 8, 2005 4:17 pm
... I intended to get the size of the complete DLE, i.e. the sum of all chunks (minus the 32K overhead of each chunk, except the first one, which turn into the...
4585
Jon LaBadie
jon@...
Sep 8, 2005 11:38 pm
A recent thread on amanda-users again raised the issue of amanda's handling of tape 'EOF', determining whether it was a tape error or the end of the media, and...
4586
Jon LaBadie
jon@...
Sep 9, 2005 2:58 pm
... What is the interaction between autoflush and taperflush? To my recollection, autoflush affects taper's actions at the start of, and perhaps during, an...
4587
Stefan G. Weichinger
sgw@...
Sep 12, 2005 7:09 pm
I once wrote about a small shellscript that I use to back up AMANDA-metadata and have received some requests to provide it here. At last I have found some time...
4588
Joe Rhett
jrhett@...
Sep 13, 2005 8:28 am
... Very good ;-) ... Difficult perhaps. But amanda is the *only* backup program that doesn't do this. ... Obvious thoughts 1. Look at what all the other open...
4589
Tobias Weingartner
weingart@...
Sep 13, 2005 3:01 pm
... Ok, I think we need to look at how some tapes are manufactured. I *seriously* doubt that each tape has exactly the same length of tape material (you know,...
4590
Jon LaBadie
jon@...
Sep 14, 2005 2:59 pm
On the users list, Guy Dallaire has been asking about his amreport that shows lots of dumper idle time attributed to "no-hold". Like 95% of the time dumpers...
4591
Scott R. Burns
Scott.Burns@...
Sep 15, 2005 2:11 am
Here are some notes on the subject I found on a SunOS 5.8 man page on the "st" driver in case this if of any value. EOT Handling The Emulex drives have only a...
4592
Ian Turner
ian@...
Sep 15, 2005 2:45 am
From a theoretical standpoint, this is an easy problem to solve: You try to re-read the last thing you wrote, and if you get it, then everything is fine. If...
4593
Scott R. Burns
Scott.Burns@...
Sep 15, 2005 3:26 am
I will prefix this with the fact that I retired from 'C' coding a couple of years ago and I only took about 1/2 an hour to look at these tools. I may be way...
4594
Ian Turner
ian@...
Sep 15, 2005 5:02 am
... On some platforms, and on some hardware. But for some combinations of platform and hardware, there is literally no way to tell the difference. ... Spanning...
4595
Ian Turner
ian@...
Sep 15, 2005 5:03 am
... On some platforms, and on some hardware. But for some combinations of platform and hardware, there is literally no way to tell the difference. ... Spanning...
4596
Jean-Louis Martineau
martinea@...
Sep 15, 2005 11:07 am
Hi Geert, I commited your patches. I also patched amanda so that the failed dump are not added to the history file. Thanks Jean-Louis ... -- Jean-Louis...
4597
Tobias Weingartner
weingart@...
Sep 15, 2005 1:19 pm
... Unfortunately, you can't do that. Tape drives cache things too... The only way to ensure that something *is* on tape is to write an EOF mark to the tape....
4598
Jon LaBadie
jon@...
Sep 15, 2005 1:24 pm
... That jogged some memory cells of old discussions dealing with the backup and read idea. The concern was whether you can just match the last block on tape ...