Am 14.12.20 um 13:57 schrieb Gene Heskett:
1. take your Replyto: back out of your headers so I can reply to the list. To reply to the list I have to do a reply all, then swap the To: and CC: lines in kmail. Or just delete your .at address, so only the list gets a copy.
Oh. Sorry. Better now? Related to "identity" in Thunderbird etc
2. my scripts are exclusively bash. I use no perl, but amanda does. So IMO a perl update ought to have broken every amanda that is being kept up to date.
perl-5.24 isn't actually up to date .. in general I think it's good to also notice these perl-warnings.
And sure, these are issues for the hidden and silent *developers* of Amanda.
3. it seems to only happen when it runs out of days to do the biggest DLE's which runs over to the second tape, so as an experiment I doubled the size of a vtape 2 days ago and changed the priority order from a string of BT's to a string of SBT's. That does not seem to have effected the average backup size. But it also has not yet hit again, the end of a runcycle where it has to do a whole pile of big DLE's resulting in the use of two tapes. In this case the crc's in the holding disk differ and it fails when using the second tape. My script needs help in that event as I wrote it in 2004 and haven't touched it since, so I'm trying to run down the conditions which trigger the fault. The finger of meager evidence I have is that it seems to occur on 2 tape nights. So now we wait for what would have been a two tape night.
The original topic of this thread was an issue happening at recovery time. Your issue happens at dump time. What they might have in common is using 2 or more tapes, so yes, maybe a similar underlying problem.
Take care Stefan, and stay well. Betsol is so far, all hat and no cattle, and this users list needs all the help it can get. I thank you for the suggestion, a lot.
It was only a basic observation and hint, you're welcome.
