Jump to content

DavidHertzberg

Members
  • Content count

    1,038
  • Joined

  • Last visited

  • Days Won

    38

Posts posted by DavidHertzberg


  1. After 3 tries, DovidBenAvraham has created a sufficiently neutral-sounding Request for Comment section on the "Backup" article's Talk page that several other Wikipedia editors have commented.  All of them so far, plus DBA, have reluctantly voted to split off the "Enterprise client-server backup" section into a separate article.  DBA's justification is that only doing the split-off will allow him to file an Administrators' Noticeboard request to ban Pi314m from editing the new separate article.  Pi314m promised in his own voting statement to act as if such a ban is in effect, but DBA pointed out that the promise is worded so as to likely expire within a day.

    Anybody who is already a Wikipedia editor can come over to the Request for Comment section on the "Backup" article's Talk page, in order to vote on and/or discuss DBA's request.  Actually you don't already need to be a WP editor, but by the time you learn to use the idiosyncratic WP editing system DBA will have already filed his Administrators' Noticeboard request—so your input may not count for much.


  2. zz-pdb,

    Both possible versions of your problem have been discussed in a Mac 9+ Forums thread.  Here's my post that distinguishes both versions; you can look forward or backward in that thread to find a possible solution for your particular version.

    However if it is indeed your "client" machine that has two network interfaces, this Knowledge Base article doesn't say how to specify the "ipsave" command for the Linux Client software.  Page 554 of the Retrospect Windows 16 User's Guide says:

    Quote

    Note:In addition to the Java-based graphical user interface, Linux clients can also be controlled through the command line. To see the command line arguments, enter the following:$retrocpl --help

    If that doesn't help, here's why and how to submit a Support Request for a bug.

    P.S.: I had forgotten to suggest that you Add Source Directly, specifying its fixed IP address on your LAN.  I see that this option in adding "client" machines, which solved  a two-year-old problem I had with -530 errors, is called Direct Access Method on pages 294 and 296-297 of the Retrospect Windows UG.  Since Subnet Broadcast—which also to some extent bypasses Multicast—may solve your problem if it is your "backup server" that has two network interfaces, you probably should read pages 294-299 in the UG.  Either way you'll need to Remove and re-Add your "client" machine.


  3. Thank you, CherylB.   Would you believe that same article, titled "macOS Mojave – Application Data Privacy",  has been in the Knowledge Base under "Top Articles" since 16 October 2018?    Great article titling, Retrospect engineers! :rolleyes:

    P.S.: CherylB, how about your creating a Support Case suggesting that Retrospect Inc. change the KB article's title to something like "macOS Mojave – Application Data Privacy  and Error -1101"?   Here's why and how to do  that.


  4. I stopped the "sacrificial" script early 8 June morning after 19 minutes when it issued  "!Error reading snapshot record ...", which it shouldn't have been doing because the script specified Recycle.  I did a Rebuild of "Media Set White" from portable HDD "G-DRIVE White" , and the "real script started to run OK—until it failed with a -519 (network communication failed) error after 4 minutes when I absent-mindedly shut down my MacBook Pro "client" preparatory to going back to bed.


  5. After I downloaded Retrospect Mac 16.1.2.102, both the Recycle backup scripts—the "sacrificial" one and the "real" one—started to run OK early 1 June morning to "Media Set Red".  However, after the "sacrificial" script ran OK, two things happened while the "real" script was running:

    • The backup of my MacBook Pro "client" failed after 1.5 hours with -559 (network connection timeout).  The script continued with my local drives.
    • After the script backed up and compared my late friend's local "Macintosh HD" and backed up my local "Macintosh HD SSD", it failed with -116 (volume doesn't exist) when I gently (I thought) picked up "G-DRIVE Blue" to look at its bottom masking-tape label one minute into the compare.

     


  6. I originally posted this on 25 May (as shown in the right-most column of the thread listing), but it was deleted by Retrospect Inc.—evidently sometime after 6 June 2019 03:35 when I copied it as an Additional Note to Support Case #67777—because they considered a final paragraph I've now omitted to be overly critical of Retrospect Inc. engineers.  This happens, as I warned you in the third paragraph here!

    Both the Recycle backup scripts—the "sacrificial" one and the "real" one—ran OK early 25 May morning to "Media Set Blue". Of the possibilities discussed here up-thread:

    1. I simply pushed the USB3 cable firmly into "G-DRIVE Blue" late 24 May morning; I didn't take special precautions. This wasn't the week for my cleaning lady, but last week—when the scripts also all ran successfully—was.
    2. The USB3 cable I used was the same one that was used when I had problems two Saturdays ago.
    3. I had upgraded my "backup server" to Retrospect Mac 16.1.0 before last Saturday's run, and I upgraded it to 16.1.1 on Thursday.

    So possibility 3 is the winner. The Retrospect Inc. engineers must have fixed the Recycle bug in 16.1.0, even though there's no mention of that in the cumulative Release Notes. I can't believe they released 16.0 without having tested Recycle for APFS-built Media Sets; I too would be ashamed! :angry:


  7. zz-pdb,

    I sincerely doubt that it wouldn't work on AMD processors, but I'm a Retrospect Mac administrator—and all Macs manufactured since 2006 have Intel processors (before that they had IBM PowerPC or Motorola processors)..

    Here's why and how to file a documentation request.

    P.S.: Here's the most recent post (from 2014) I could find where the OP says he is running Retrospect on an AMD CPU.  Here's an 8-month-old post where the OP is having the same problem as jgaiche reported in 2014  but is running Retrospect Windows 15.6 on an Intel CPU.  FYI, here's a post from farther down x509's 2018 thread where he recounts Retrospect Tech Support's response to that same problem, which AFAICT will only be resolved for Retrospect Windows administrators by the introduction of the glorious Web-based Management Console.


  8. I've noticed a recent rapid uptick in the number of views of this thread. I'm guessing that's because a number of you have noticed drastic changes to the ""Backup"" article, including deletion or total-dumbing-down of two paragraphs in the "Performance" sub-section of its "Enterprise client-server backup" section (described here).  DovidBenAvraham is now trying to remove the article from the clutches of what he and I would describe as a "neatnik near-nut-case".  If any of you are Wikipedia editors, please don't interfere at this stage—which consists of getting a Third Opinion from a WP outside volunteer.

    To quote DBA [edited slightly]:

    Quote

    The overall picture that emerges [in immediately-preceding paragraphs by DBA on the article's Talk page] is of Pi314m [the "handle" of the other editor] deciding without any discussion to consolidate a whole series of related articles into a single article that conforms to his concept of the subject matter. He can technologically get away with this flouting of Wikipedia rules, apparently because because he is doing a copy-paste of the merged-in article's text followed by replacing that text with a redirect to the moved-to article. The only reason I [DBA] caught Pi314m is because he did copy-pastes between a sub-section of the "Enterprise client-server backup" section and preceding sections of the same Backup article. As is his custom, he did not discuss these "merges" on that article's Talk page; if he had, I [DBA] would have carefully explained (as I [DBA] now have on Pi314m's personal Talk page as well as in the [article Talk page] section above this) that application feature descriptions in the last section of that article may seem like duplicates of the same-named feature descriptions in preceding sections of the article—but they're not.

    IM[DBA's]HO the underlying problem is that—as I {DBA] have shown in the preceding [article Talk page] paragraphs—Pi314m believes that Wikipedia gives him the right to be the sole decider of the subject and contents of an article, even when it's partly about subject matter of which he knows nothing. That would explain why he has not responded to my [DBA's] subject-related comments in the [article Talk page] section above, and why he reverted my [DBA's] 26 May edits that put back the two feature-description paragraphs he had deleted from the "Performance" subsection of the "Enterprise client-server backup" section; he considers that I [DBA] have no right to second-guess his decisions.

     


  9. Thank you, Nigel Smith, for suggesting a problem explanation—launchd—I didn't know anything about.  minidomo, did you inherit and/or upgrade this Retrospect installation?

    minidomo, read all pages 198-204 in the Retrospect Mac 15 User's Guide.  "launchctl on the Mac" on pages 203-204 discusses what Nigel has diagnosed, with a slightly different (but possibly out-of-date, because the chapters after "What's New" in the UG haven't been updated in years) solution.


  10. x509,

    Look at pages 362-367 in the Retrospect Windows 16 User's Guide, and see if you can figure that out.

    You might have to check-mark the Set Archive Attributes option per page 364, Restore those photo files, and change the Windows date values for them again to cause the files to be backed up again.  Then again, I might not know what I'm talking about. B)

    P.S.: Wait a second, aren't those file date values stored in the Snapshot entry for a file?  If so, then Retrospect would have updated all of the entries on the next Backup run.  I know that because, as a holdover from when I was having a problem with -530 errors, I  schedule a "sacrificial" script run before each "real" Backup script run.  The "sacrificial" scripts specify the No Files Selector (Rule for Retrospect Mac)—meaning no files are actually backed up, but they spend about 3 minutes "creating snapshot" and "copying snapshot"—evidently for each file on my MacBook Pro's SSD.  And yes, "Windows Selector Conditions" on page 440 of the UG says:

    Quote

    Date (Windows) uses Windows-specific date attributes as conditions.

    P.P.S: For Retrospect Mac administrators who may be reading this, Snapshots are the "crazy Mrs. Rochester in the attic we're not supposed to talk about".  Back around 2000, first Unix-based OSes and then Windows developed this capability which the industry has termed a "snapshot".  Thus when Retrospect Mac 8 was developed in 2008, EMC documentors decided to drop the use of the term "Snapshot" for a  different capability that had been in Retrospect since the late 1980s; however the terminology in Retrospect Windows has never been changed due to "unfortunate events" in 2009-2010.  DovidBenAvraham used to have this described in the Wikipedia article, but other meanie WP editors made him remove this as Original Research.  So you'll have to sneak over and read about Retrospect Snapshots on page 31 of the Retrospect Windows 16 User's Guide.  But don't talk about them, even though you can hear "Mrs. Rochester screaming" every day in messages issued by the Retrospect common underlying code.;)  "Mrs. Rochester" doesn't "burn the house down" as in Jane Eyre; in fact Snapshots are a vital part of Retrospect Mac as well as Retrospect Windows operations.  Pages 229-230 of the Retrospect Mac 16 UG say "Retrospect now uses the term backup to include both session and Snapshot data." B)  BTW, APFS offers "snapshots" in the industry-standard sense, so the EMC documentors were prescient.


  11. minidomo,

    Am I correct in thinking that the messages you refer to in your third paragraph immediately above are in the macOS Console log, not the Retrospect log?  [assumes Indian accent] You should PM me your credit card information, so that our technical specialist can remove a very serious virus infection from your Mac! [ drops Indian accent] ;)  Seriously, maybe you should stop looking at your macOS Console log.  I assume you've checked your Retrospect scripts to make sure that "Fridge" and/or its folders are not check-marked as Sources or used as Destinations in any of them.

    My guess is that you're getting those messages either because you're still running Time Machine, or at because at one time you specified a Time Machine Backup Disk.  If it's the latter, you should go to System Preferences->Time Machine, click Select Backup Disk, click the "Fridge" disk that's shown there, and then click the Remove button to remove it.  If it's the former, then you should know that Time Machine doesn't work with an APFS-formatted Backup Disk (which is what the messages are telling you)—so reformat "Fridge" with HFS+ (which you may not be able to do if it's an SSD).

    If you're not running macOS Server or Windows Server on any of your machines, then Desktop Edition is all you need.  As I've delicately alluded to in this post and the following one in another thread,  Retrospect Inc. is worried about reduced Edition license revenue because many installations have substituted Linux-based file servers for macOS Server—which doesn't provide much anymore beyond what you get for free with macOS 10.14.  An upgrade to Retrospect Mac 16 Desktop Edition will cost you US$70-75. :)


  12. minidomo,

    I don't know know what the cause of this specific error is, but I'm pretty sure that "retroisa" is Retrospect's separate Instant Scan process.  If you un-check the Enable Instant Scan box per page 200 in the Retrospect Mac 15 User's Guide, that should eliminate the repetitive Console error message.  The cost will be some slowness in your backups. 

    BTW, is this message in the Retrospect Console log, or in the macOS Console log?

    Another thought that occurs to me is that your Mac Mini may have a drive formatted with APFS, especially if it's running macOS 10.14 and the drive is an SSD.  Several administrators found that Retrospect Mac 15 was taking hours to scan APFS-formatted source drives, and a Knowledge Base article was updated in May 2018—when Retrospect 15 was the latest major version—to state that Retrospect Instant Scan doesn't work for APFS.  As stated in this final post in that same thread,  after my MacBook Pro "client" SSD was forcibly converted to APFS I upgraded to Retrospect Mac 16; I found that the MBP scan is taking less time without Instant Scan than it did under Retrospect Mac 15—when the SSD was still formatted with HFS+ and I was using Instant Scan.  (I should point out that Retrospect Inc. and its predecessor corporations have never paid me a cent, so I won't get a commission if you upgrade to Retrospect Mac 16—and you may be able to get away with upgrading to the Desktop Edition because I don't think Retrospect can identify recent editions of macOS Server B) .)


  13. Fredspon,

    After rereading this, I question whether you should set your Scheduled Executions as Recycle.  You say "I want the original backups to remain intact as they are recycled by their own scripts after 3 months".  However page 226 of the Retrospect Windows 15 User's Guide says:

    Quote

    Recycle clears the Catalog contents (if any) of a Backup Set so it appears no files are backed up. Then it looks for the first media member of the Backup Set and erases it if it is available.

    If I understand what you probably want to do, it is to Recycle your primary Backup Set disk Members after 3 months to prevent their running out of space, but to keep accumulating forever the backups on your secondary disks in case you need those backups for disaster recovery.  This is the user-demanded feature that the EMC predecessor of Retrospect Inc. built with Transfer in the early 2000s; many enterprise users wanted—and still want—their secondary Backup Set Members to be on tape (beats chest, shouts "Real enterprises use tape!" ;)) because buying an additional high-capacity blank tape every so often is cheaper (once they've shelled out upwards of US$1500 for a tape drive) than buying an additional hard disk drive every so often.

    If this is what you want to do, then—besides changing Scheduled Executions to Normal—you should tick Match Source Volumes to Catalog File and tick Don’t Add Duplicates to Backup Set and tick Transfer Any Needed Intermediate Database Snapshots (I've initially-upper-cased some more words in these option names to make the names easier to read).  Although those options will slightly increase the running time for your Transfer Backup Set scripts, it will allow you to Transfer incremental backups while preventing any multiple transfer of the same backup.  You won't have any unpleasant surprises after 3 months, and I doubt you tested for that long.

    I should point out that I am a lowly home user of Retrospect Macintosh, who does a Recycle backup every Saturday onto a brought-back-from-bank-safety-deposit-box portable HDD and does incremental backups onto it Sundays through Fridays.  So an administrator with real enterprise experience should correct me if I'm wrong. :rolleyes:  However I just found my applicable three-year-old post, and the options in it are the same ones I said to tick in my third  paragraph (second below the quote) of this post.  I'm not providing a link, because the thread it's in is about doing Cloud backups using Retrospect Mac and involves initial "seeding".  However three years ago I changed the options ticked in the post after the head of Retrospect Tech Support posted a Tutorial, to which I also won't link because I see the Transfer options in Retrospect Windows have changed since the Tutorial was posted.

    P.S.: If what I suggested in the paragraph below the quote is not what you want to do, at least have your Transfer Backups Sets script(s) be scheduled to do a Recycle at least one run after your Backup script(s) are scheduled to do a Recycle.  That way you don't run into the following  bad situation, which Murphy's Law says will happen:

    1. One or more data folders on your source drive go bad.
    2. Your Backup script runs with a scheduled Recycle immediately after that happens, so your primary Backup Set no longer contains data from the now-bad folder(s) .
    3. You don't notice the error message(s) from your Backup script, so you allow your Transfer Backups Sets script to run with a scheduled Recycle.
    4. Result is that you no longer have that data on either your source drive or your primary Backup disk Member or your secondary Transfer Backups Sets disk Member.  Back in the early 2000s your Transfer Backup Sets script(s) would have had tape Members as their destinations, so (provided you hadn't re-used your existing tapes for the Transfer Backups Sets run that did the Recycle) you could have done a Recreate of your secondary Backup Set's Catalog file from the tape Members and done a Restore using that secondary Backup Set; however you're probably not using tape.

  14. cmgsupport and PaulMikeC,

    This Knowledge Base article says:

    Quote

    Retrospect triggers these events by calling a file named [my emphasis] "RetroEventHandler". This file can be any script type known to the operating system (As examples for Windows, BAT or VBScript. For Mac, bash or ruby).

    My eventually-achieved understanding of the first sentence in that quoted passage is that you have to rename your script file to "RetroEventHandler" (without the quotes).  It is not sufficient to put the script file in the proper folder per the KB article.  Did you do the renaming?

    Sorry it took me so long to spot that; I looked at the KB article several times. :rolleyes:  I haven't messed with script files myself, but you could PM saskia.

    P.S.: In view of what PaulMikeC has written directly below, my apology to him—and presumably to cmgsupport—for suggesting that he missed the KB article statement that the script must be named "RetroEventHandler" .  However I strongly recommend that he file a Support Case suggesting that either the KB article or the sample event-handler script(s)—or both—be revised to deal with the fact that the log file isn't generated in a directory where the script creator would expect to find it.  Here's why and how to do this.


  15. Indi-Tech,

    First, I think you got a bit too cute with your screenshots.  Favorite Folders belong to drives, but you edited-out the parent drive(s)with its/their disclosure triangle—although I think I see the ghost of one at the very top of your topmost "After browsing" screenshot.  Also, assuming these are screenshots of the Sources panel, you appear to have re-arranged the columns so that the relevant ones would appear within the width of a Web-browser window.

    Second, I did a Forums search; the only recent part-of-a-thread that seems relevant starts with this post.  Although I've been using Retrospect Mac for 19 out of the last 24 years, I only started using a Favorite Folder about 6 months ago.  But that Favorite Folder, with which I've had no trouble, is on a HDD that is locally attached to my "backup server"—whereas you state that that these Favorite Folders are on drive(s) on a "client" machine.

    Based on what henry-in-florida said a couple of posts above the post linked-to in the preceding paragraph, I'm now going to make a wild-a**ed guess that your problem is somewhat similar to the problems I had for two years with -530 Client Not Found errors—and that these Favorite Folders are the on the first "client" backed up in one or more of your scripts.  The short  version of my cure would be to Remove the "client" in Sources and re-Add it with Add Source Directlynot Use Multicast.  Before doing this you'll have to make sure that the "client" has a fixed IP address on your LAN; after doing this you'll have to re-checkmark that "client" and its Favorite Folders in each of your applicable Scripts, dragging the "client" in the Summary list pane of each Script in order to get it into the desired backup sequence.

    I actually got the clue for the cure for my -530 errors from something henry-in-florida wrote.  The explanation I eventually came up with for my -530 problems is described in the last quote and the two sentences following it in this post in another thread.  The symptoms you describe for your -1101 errors sound similar to my Phase I -530 situation starting in February 2017.  My first cure, which worked until November 2017, was to schedule a "sacrificial" script 5 minutes before each "real" backup script, in which only the first "client" was specified and in which the No Files Rule was used.  My assumption, based on a suggestion made by a networking expert on the Ars Technica forums, was that my -530 errors were caused by a timing error during startup of the "backup server" Engine—since I was only getting -530 errors if I didn't boot my "client" and "backup server" machines until at least 10 minutes after the "real" script's scheduled start time.  After this stopped working in November 2017, I found that I could avoid problems by pausing all scripts, clicking  the "backup server" Locate button for my first "client", and then killing the "sacrificial" script.  That sounds rather like what you are doing to temporarily avoid -1101 errors.

    Here is how and why to submit a Support Case.  You'll notice that Retrospect Inc. personnel no longer look at the Forums, even the "Retrospect bug reports" one.  The Retrospect Inc. engineers did try to find the cause of my -530 bug in May 2018—many months after I submitted a Support Case, but their test version of the Engine and Console didn't fix it.  My guess is, as reflected in the first quote in this same post in another thread, is that Use Multicast—which once worked beautifully in Retrospect—is now inhibited by features added to some home networking equipment.  The engineers in Walnut Creek CA  simply don't have the equipment to reproduce the problem, so they have tried to find it from logs etc. they requested I submit.

    P.S.: henry-in-florida reported that Adding with Subnet Broadcast worked for him, and he didn't need to Add Source Directly; I tried it briefly last year and it solved my -530 problem.  If I'm reading page 79 of the Retrospect Mac 16 User's Guide correctly, Subnet Broadcast doesn't use the now-troubled Multicast feature.  Like Use Multicast it allows a backup administrator without IT training to add additional "clients", but only after an IT-trained person—IMHO one with even more knowledge than is required to set up fixed IP addresses—has set up Retrospect.

    P.P.S.: Indi-Tech, if you're going to us these Forums for help you should check recent threads before starting a new one.  twickland reported the same problem with Favorite Folders less than a month ago; for some reason his post didn't show up on my Forums search.  He was using Retrospect Mac 16.0.1; does your "Latest version of Retrospect" mean that same version?  There is now a 16.1.1.100, BTW.


  16. 1 hour ago, Indi-Tech said:

    We are an independent technology consulting and support company.

    Counter to your statement, we are using Retrospect at the low end of the scale. Our own knowledge base of Retrospect support tickets goes back to 1999, 2 years older than the Apple ID of the company founder was established. 😎 The big boss does recollect something about this issue in the deep past and having to, "bind",  the IP in a sort of way. Not because of multiple NICs but because of scanning subnets and having issues, or something like that.

    Our content creation clients with 100TB plus RAID units from Facilis and 45 Drives could never use something like Retrospect and have files backed up in time. At least, not based off our tests. We use Retrospect for office operations and for home professionals, small businesses with smaller files. You know, like it's the 1990s again. 😎

    We recently had a working Retrospect backup setup mysteriously forget the favorites that were set, preventing the backup from working until the favorites were re-established. No change to the setup was made before the malfunction occurred. Not the kind of reliability we can use with our bread-and-butter cliental. And exactly the kind of behavior Retrospect has always exhibited. It's a good product but seems to have been stuck in time in some aspects.

    Gee, Indi-Tech, and yet you only joined these Forums on 2 May 2019.  ;)  Maybe your big boss belonged under a different Forums "handle"?  Anyway, good of you to condescend to ask for help from us home professionals. :rolleyes:

    I'll deal with your Favorite Folder problem later, to the extent I can, after I've had dinner and watched a movie.


  17. On 5/21/2019 at 12:54 PM, Nigel Smith said:

    I've not seen it in the manual but, to be fair, dual active NICs on a client isn't common and they can't document *everything*. Searching the KB for "bind" gets me that linked page as second hit -- although you have to be techie enough to know the term "bind" in relation to NICs and IPs...

    You don't always have to remove and re-add the client -- depends on how it was added in the first place. But it won't hurt to do so unless you are trying to preserve an easily-followed backup history (and, usually, you do this kind of shenanigans because you can't get any backup to begin a history with!). 

    Connection not showing might be a client-version thing -- certainly works for me in e.g. 15.6.0

    As to speed -- unencrypted will be faster than encrypted, fewer large files will be faster than a lot of smaller ones,  and so on. Try your own tests and see where the bottleneck is, watch the network usage, CPU %s, and disk I/O rates at both ends, etc. And when you say encryption are you talking encrypted communications, encrypted backup set, or both? (IIRC, Chronosync offers encrypted comms but not the backup itself, instead recommending you use an encrypted disk if you want that sort of thing. But it's been a while...)

    Indi-Tech,

    Retrospect Inc. could have, in fact, put a sentence or two about "ipsave" in the User's Guide.  However a search of the cumulative Release Notes shows this feature was added in September 2014 to Retrospect Mac, and in March 2015 to Retrospect Windows.  It therefore ran afoul of the User's Guide policy described in this post in another thread (which is not the only place I have complained about it).  But Google shows you're a consultant (or someone has stolen your "handle" and logo), so you'll have to develop an encyclopedic mastery of the Knowledge Base articles to justify your charging the big bucks—which is why I think Retrospect Product Management intentionally adopted this "disappearing documentation" policy as part of its "go big or go home" strategy. ;)

    lwhitten complained about the connection not showing in this post I linked to above.  He/she was using "a trial of [Mac] v14", and you're using Mac version 16, so Nigel Smith must be somehow hitting the sweet spot in Retrospect versions—or there's another explanation.

    As for the speed of backup over a LAN, dittberner@dbr3.de complained about that for nearly 2 years before switching to another client-server backup application.  I came up with a hypothesis here, but my hope—expressed a couple of posts down in that thread—that things would improve with version 16 has not been realized. :(  My average speed in Recycle backups over my home LAN is around 600MB/min; my hypothesis might explain why ChronoSync—which you're probably using in Synchronize mode—is faster.

     


  18. As you can see from the color-named progression ("three cheers for the Red, White, and Blue") of my Media Sets in the OP, this morning I was scheduled to do my Saturday Recycle backup to "Media Set White"—whose "1-Media Set White" Member is on portable HDD "G-DRIVE White".  Since the Retrospect Agent Reply in the post directly above this evinces a certain skepticism about whether my portable HDDs have been properly cabled to my Mac Pro "backup server" when I switched them, after I had yesterday un-cabled "G-DRIVE Red" (to take it to my bank safe deposit box) and cabled "G-DRIVE White"  I ran a  "sacrificial" script with No Media Action backing up my MacBook Pro's SSD to "1-Media Set White".  This ran OK in 4 minutes. I then scheduled a repeat of the same "sacrificial" script to run 5 minutes before the "sacrificial" script with Recycle backing up to "1-Media Set White" early this morning.

    All three scripts ran OK in succession :) early this morning backing up to "1-Media Set White" : the No Media Action "sacrificial" script backing up my MBP ,  followed by the Recycle "sacrificial" script backing up my MBP, followed by the Recycle "real" script backing up my 4 drives—starting with my MBP's SSD—and a Favorite Folder.  I have to admit I was disappointed with missing the opportunity to demonstrate my bug-proving prowess to Retrospect Tech Support. :rolleyes:

    I can think of only 3 possibilities for why the Saturday scripts ran OK this morning:

    1. My "G-DRIVE"'s cables had previously been jiggled loose by me or my cleaning lady on previous Fridays, but not yesterday.  This seems unlikely, because I'm pretty careful and because my cleaning lady only comes on alternate Fridays.
    2. One of my two "backup server" USB3 cables, which are swapped into use every Friday morning, is defective.  This seems unlikely, both because the OP shows failures on four successive Saturdays and because the same two cables worked fine connecting "G-DRIVE"' portable HDDs on Sundays through Fridays.
    3. On Friday morning I upgraded to Retrospect Mac 16.1 from 16.0.  Whoever wrote the Retrospect Agent Reply in the post directly above this made that upgrade suggestion; maybe he/she knew knew something about an undocumented bug fix involving Recycle and APFS.

    I'll test again next Saturday morning.

     

     


  19. Retrospect 16.1 for Mac and Windows has "New Retrospect Management Console: Pause/Unpause/Stop support", per this post in another Forum.  Is that a truly new capability for a Web-connected Retrospect monitoring application (as opposed to the LAN-only Retrospect Mac Console), or is it already in Retrospect for iOS?

    If the capability is already in Retrospect for iOS, please post on what precisely the capability is.

    P.S.: I just found this blog site, with posts claiming to have developed Retrospect for iOS up through version 2.  So Brian Dunagan, if you're not too busy being COO of Retrospect Inc., maybe you could obtain from one of your subordinates an answer to the above question.:)

     

     


  20. Despite the lack of non-Partner-applicable progress (other than "Pause/Unpause/Stop support", which I think may be already in Retrospect for iOS) described in the last paragraph of this post in another Forum, DovidBenAvraham will try to keep the mentions of Shared Scripts here and here in the Wikipedia article.  If the promised deployment enhancements aren't in Retrospect 16.5—expected in early September 2019—however, DBA's afraid those mentions are going to have to be removed. :(


  21. Quote
    
    Retrospect Agent Reply can be found below. 
    
    Agent Response: 
    The issue is this specific error: 
    !Backup Set member "1-Media Set Red": error -1,102 (drive missing/unavailable) 
    !Can't reset Backup Set Media Set Red, error -1,102 (drive missing/unavailable) 
    It looks like Retrospect could not find this media during a recycle backup attempt. 
    Do you know the status of the media when this error was reported? 
    Retrospect version 16.1 has been released, please update to the latest version.
    

    To which I replied [start-end underscored words changed to underlined words, which the Forums allow but the Support Case software doesn't]:

    "G-DRIVE Red", which is a USB2 portable HDD, was—and still is—cabled to my Mac Pro "backup server". Note that, as soon as the "sacrificial" script terminated on 11 May with the -1102 error. the "real" script began to run. It proceeded several minutes after logging "Finished scanning backup set data files" before I stopped it, which I did because the Summary was showing that the run was methodically examining the existing files in "Media Set Red"—despite all 3 schedules for the script specifying Recycle. Then, after I did a Rebuild of "Media Set Red"—which I stopped after the log said "Adding to 29.9 GB of backup set data, starting at file index 51", I'm pretty sure I Finder-deleted file "1-Media Set Red" on "G-DRIVE Red" and Removed/reAdded "Media Set Red"—which doesn't show in the log. After that a Repeat Never schedule of the same "real" script to "Media Set Red" ran OK.

    Note that the same sequence of script runs the previous Saturday 4 May to "Media Set Blue", with its member on "G-DRIVE Blue", got pretty much the same results. The only differences are that I manually-scheduled a run of the "real" script I created after stopping the automatically-scheduled run, and stopped that run after it logged "!Cannot find parent directory for file Scalloped-Frame-V-2.png in backup 4/13/19 4:06:29 AM"—which certainly implies that it wasn't doing a Recycle, and that I let the subsequent Rebuild for the existing "1-Media Set Blue" run over an hour to completion.

    I alternate portable HDDs weekly between 2 USB-A-to-USB-Mini cables attached to different built-in ports on the back of my Mac Pro "backup server". Note also that these 3 portable HDDs always work fine for the rest of the week's No Media Action scheduled runs. So, despite my not having attended Hogwarts, I think there's a bug in Retrospect Mac 16.0 for Recycle.

    I will now proceed to upgrade to Retrospect Mac 16.1 for both Engine-Console and Client.


  22. It being the evening of 14 May—close to my expected date for the release of Retrospect 16.1, ""I dipped into the future" again.  The Knowledge Base articles haven't been updated yet—with the exception of "How to Set Up the Management Console", but the 16.1 downloads are on the website.  As for updated "What's New" chapters in User's Guides, I don't want to laugh so soon after dinner. ;)

    The question applicable to this thread is "What improvements have they made in the Web-based Management Console?"  As shown by the 16.1 version of the cumulative Retrospect Windows Release Notes, the answer appears to be "not too many, but promising".  For the Engine they are:

    • New Retrospect Management Console: Pause/Unpause/Stop support
    • New Retrospect Management Console: Versions, Editions, Platforms listed
    • Improved Retrospect Management Console: Disable deployment for an existing shared script
    • Improved Email Notifications: stopped scripts now generate email with title "Execution stopped by operator - Retrospect"

    For both Engine and Windows Client there is also—and this will make a lot of administrators happy :) :

    • New Windows 10 May 2019 Update certification

    The first and third Engine improvements actually IMHO sound like a real step forward in 2-way capability. :)  However as of this evening it's evidently not enough to persuade Retrospect Product Management to actually insert the Management Console Add-On into the Product Configurator, so that ordinary non-Partner administrators can add 2-way Console capability (for US$49 or more) to their installation of Retrospect without saying "pretty please with cherries on top" ;) to a salesperson on the other end of a phone line.  How the dickens do they expect the Management Console's 2-way capability to be adequately beta-tested without the help of ordinary administrators? :rolleyes: (And testing, for the few of you who don't already know, has been Retrospect's product weak spot for the past 30 years. :()

    OTOH there appears to be no progress towards the "Management Console - How to Use Shared Scripts" KB article promise that "Support for local sources, local destinations including disk, scheduled scripts, and more extensive scheduling options will be available soon". :(  I have a jaundiced vision of running battles in the halls of Walnut Creek between developers shouting "Make the GUI look like Retrospect Windows!" and other developers shouting "Make the GUI look like Retrospect Mac!"  I hope there isn't blood spilled. B)  However a more developer-realistic vision is that the engineers just didn't have time to make releasable progress on that front, because they were fixing 9 bugs—of which 4 for Retrospect Mac (only 2 of which affected Retrospect Windows) involved Storage Groups.

    P.S.: There is progress as of late October 2019 on "Support for local sources, local destinations including disk, scheduled scripts, and more extensive scheduling options".  See this post in a more-recent thread, and the Granular Remote Management webinar it links to in the third paragraph.


  23. Thank you, Nigel Smith, for that suggestion. :)  I'd forgotten about this prior thread of Nigel's, to which I made a few ignorant contributions.  However it looks as if this post in an earlier thread is closer to solving Indi-Tech's problem.

    Nigel's problem dealt with "clients" that can be on either of two different subnets when they connect to his LAN, with his "backup server" machine only sitting on one of the two subnets.  lwhitten's problem, in the earlier thread, was that—for speed in backup—he/she wanted his "client"'s Client to access his "backup server" over a dedicated cat6 cable port rather than via a router with dynamic DHCP IP addresses.  Using ipsave—described in the "Binding Retrospect Client to a specific IP address" Knowledge Base article Nigel linked to above, lwhitten found it worked OK for scheduled backups and "kind of OK" for Proactive backups.   Indi-Tech's "LAN data server" that is also a Retrospect "client" is, I confidently assume, not being backed up with a Proactive script—since it is undoubtedly always connected to the LAN.

    In case it's not obvious why I'm making this post, it's so other administrators will in the future be able to find a solution for similar problems by doing a Forums search.  An enabling Client "ipsave switch" was fixed in Retrospect Mac 11.5 and in Retrospect Widows 10.0, but Retrospect Inc. never described its function in either variant of the User's Guide—only in the KB article.


  24. 1 hour ago, Indi-Tech said:

    How about starting with the client having (2) NICs and tell us how to get the Retrospect server to connect to a specific client NIC.

    Indi-Tech,

    I have no idea, except that it sounds as if the "client having (2) NICs" is actually a LAN data server, which implies that it is shared on the LAN by the "Retrospect backup server" machine as well as other machines.  If so, is there any reason that you can't back up the "client having (2) NICs" as if at least its "server_RAID" drive is a network share to the "Retrospect backup server" machine rather than as a "client"?  See pages 71-72 of the Retrospect Mac 16 User's Guide.

    Please be aware that I'm a home user with a simple LAN, composed of "client" machines and a single "backup server" machine, all of which have only one NIC apiece, and I don't have a LAN data server.  You'd have to set up an afp: or smb: address that would somehow specify using the "LAN data server"'s secondary IP; this makes it seem as if you can do it with AFP.

    P.S.: Changed first paragraph to suggest that "server_RAID" be treated by the "Retrospect backup server" as a network share, and split off second paragraph with caution on how to do it.

×