jquiros Posted September 22, 2003 Report Share Posted September 22, 2003 Hmmm.... I spoke too soon! Now I've gotten an~ Internal consistency check failed: Assertion check at "elem.c-812" as if it matters to anyone at Dantz, who is still "working on the problem...." Link to comment Share on other sites More sharing options...
ringg Posted October 28, 2003 Report Share Posted October 28, 2003 Dantz, do you want a little more data on the "Assertion Check elem.c-817" error? I am using Retrospect Server v. 5.1.167 on a Mac G4/400, 1GB RAM running OS 10.2.8. I perform incremental backups on about 25 clients running OS9 and OSX, Windows 95, 98, NT, and 2000 on a nightly basis. The backups are scripted to start at about 10 PM onto 5 storagesets using a Quantum DLT 4500 SCSI Autoloader. I have been using this configuration for the last 3 years, except for updates, of course. Last Friday night the backups started fine, but stopped during the 4th storageset. When I came in to work on Monday morning and sat down to my computer, the backup resumed. I stopped it manually. The log said that I stopped it on Friday night. I then manually ran the script for "Storageset #1". Retrospect backed up 4 clients successfully, but stopped on the 5th, a G4, with the "Assertion Check elem.c-817" error. I rebooted the computer and ran the script again. This time the backup ran fine - exept that it did a full backup on Windows 2000 machines. On Monday night it ran all 5 scripts, and did incremental backups on all computers except for all of the Windows 2000 clients and one NT 4.0 SP6a client, for which it did Full Backups. These clients are spread out over 5 storagesets. The only recent change in my configuration was that about 3 weeks ago I added 512 MB RAM to make a total of 1GB. Since then I haven't had any problems with anything (Photoshop, Powerpoint, etc)...until yesterday. Hope this helps. Let me know when you fix the problem. Link to comment Share on other sites More sharing options...
Mayoff Posted October 28, 2003 Report Share Posted October 28, 2003 We have seen this problem specifically caused by bad RAM. I know of more then one user who has fixed the problem by replacing RAM. Link to comment Share on other sites More sharing options...
sparker Posted October 28, 2003 Report Share Posted October 28, 2003 I am seeing this problem now too running Retrospect 5.0.238, RDU 3.1.105 on a G3 running 10.2.8 with 384Mb of memory and Retrospect being the ONLY application running. I used to run the Desktop server on the TiPB and never had any problems), but decided to move it to a new G3 dedicated to being my "home server" and treating my TiPB as a lowly client.... I'm trying to backup my Ti PB running client 5.0.540 under OS 10.2.8 with 512Mb of memory (no apps running). The log file shows two errors, the first of which is unrelated to the assertion check, and according to tech support can be ignored. The pertinent section of the log file reads: Can't read file "retropds.22", error -40 (file positioning error), path: "Titanium Banana/Applications/Retrospect Client/Retrospect Client.app/Contents/Resources/retropds.22". Internal consistency check failed: Assertion check at "elem.c-817" The seemingly suggested "solution" to the second error is to "replace the memory in the machine", but I: 1) Don't feel I should need buy new memory as a solution to a software problem, 2) Have tested the memory using TechTool multiple times, and it passes every time, and 3) Have swapped the memory around in the slots, and they still pass TechTools test. Does anybody know of any other possible solutions out there, possibly including fixing the software so it doesn't fatally die from errors like this? It's interesting to note that this may be related to Mac Clients, as I have one Mac client (my TiPB) and one Windows client (a Dell C400 running Windoze 2000). The Windoze machine backups up perfectly, but every backup "set" I have defined on the Mac client has exhibited this problem. I have one set to "duplicate" my iTunes library, one set to "duplicate" my D100 photo library, and a third "normal backup" set to get everything else. All three Mac Client backup processes have shown this error, the PC one ran fine (either a fluke or an indicator of where the problem may lie) As always, looking for possile solutions to this bug. Thanks, Steve Link to comment Share on other sites More sharing options...
CallMeDave Posted October 28, 2003 Report Share Posted October 28, 2003 Quote: The pertinent section of the log file reads: Can't read file "retropds.22", error -40 (file positioning error), path: "Titanium Banana/Applications/Retrospect Client/Retrospect Client.app/Contents/Resources/retropds.22". Internal consistency check failed: Assertion check at "elem.c-817" The -40 error is unrelated; it's a bug in 5.0 that was fixed in 5.1, and occurs for all users of the older version. The assertion check error is what's relevant. And it generates its own log entry, although only a debug version of Retrospect will yield information helpful to Dantz. >The seemingly suggested "solution" to the second error is > to "replace the memory in the machine", "seemingly?" Mayoff's suggestion seemed pretty specific. >but I: >1) Don't feel I should need buy new memory as a solution to a software problem, >2) Have tested the memory using TechTool multiple times, and it passes every time, and >3) Have swapped the memory around in the slots, and they still pass TechTools test. 1) Buying new memory would be a solution to a hardware problem which is being exposed by an application that puts greater demands on memory then any other application you're running on the machine. Do you want your backup software to be forgiving? Or do you want it to be accurate... 2) Useless. Do the math; it would take eons to completely test memory the TechTool way. Note the reports of early 15" PowerBook "FW 800" users who's memory caused kernal panics, yet passed TechTool and Apple Hardware Test CD tests. 3) How would swapping faulty memory around in slots do anything to help? Perhaps if you were able to add a bunch of additional good memory to the machine you might lessen the chances of Retrospect writing to the bad memory, but that's a silly solution (since you'd be buying more ram anyway). Quote: It's interesting to note that this may be related to Mac Clients ... snip ...All three Mac Client backup processes have shown this error The retropds.22 error is a Mac client specific error. You'll always get it on Mac clients, and you'll never get it from Windows clients. Are you reporting that all of the following is true every time: - The elem.c-817 assertion check error happens only on the Mac clients - The elem.c-817 assertion check error happens every single time you get the retropds.22 error - The elem.c-817 assertion check error never happens on the Windows clients? Dave Link to comment Share on other sites More sharing options...
sparker Posted October 28, 2003 Report Share Posted October 28, 2003 Quote: CallMeDave said: The -40 error is unrelated; it's a bug in 5.0 that was fixed in 5.1, and occurs for all users of the older version. Thanks for the clarification. I got that impression from reading previous posts, but you've made it crystal clear... I'll arrange for an upgrade ASAP. Quote: CallMeDave said: 1) Buying new memory would be a solution to a hardware problem which is being exposed by an application that puts greater demands on memory then any other application you're running on the machine. Do you want your backup software to be forgiving? Or do you want it to be accurate... I want my backup software to be forgiving AND accurate. Your statement "puts greater demands on memory then [sic] any other application you're running" makes an assumption that's incorrect. One of the applications I run is Photoshop 7 which is famously memory intensive, especially when editing Nikon RAW images from a D100. I don't have problems with Photoshop (other than it's quite slow on this machine)... Quote: CallMeDave said: 3) How would swapping faulty memory around in slots do anything to help? Perhaps if you were able to add a bunch of additional good memory to the machine you might lessen the chances of Retrospect writing to the bad memory, but that's a silly solution (since you'd be buying more ram anyway). It "might" move the previously suggested problem RAM into a different location, causing the OS or other applications to fail earlier or in a more apparent way... However, every application and test I'm running shows that there is no 'faulty memory" anywhere on this machine. Quote: CallMeDave said: The retropds.22 error is a Mac client specific error. You'll always get it on Mac clients, and you'll never get it from Windows clients. Are you reporting that all of the following is true every time: - The elem.c-817 assertion check error happens only on the Mac clients - The elem.c-817 assertion check error happens every single time you get the retropds.22 error - The elem.c-817 assertion check error never happens on the Windows clients? Yes, as of this time the above three statements are true (I've only been running "Desktop" on this new machine for a couple days.) And there is one more statement that's true as well: The elem.c-817 error has occured without being preceeded by the retropds.22 error during a "Duplicate" operation (duplicating folders that don't include "Retrospect Client"). What's interesting in this case is that I can immediately (without restarting the G3) re-run the script for the duplicate job after such a failure, and it will complete normally. You mention that there's a debug version of Retrospect that can help Dantz diagnose problems. How do I get ahold of this? I'm making an assumption from the tone of your previous message that you work for Dantz. I've been a loyal customer of Dantz for a very long time now, and this is the first time I've run into problems that prevent me from performing a successful backup. Please don't get me wrong. I'm trying not to "point fingers", but want to figure out a solution that doesn't involve buying more hardware for a machine that operates perfectly in all other software that I have running on it. Steve Link to comment Share on other sites More sharing options...
sparker Posted October 29, 2003 Report Share Posted October 29, 2003 "CallMeDave", Thanks for the idea to upgrade.. I purchased my 5.1 upgrade today, installed it this afternoon, and have completed a backup of the remote client that was causing me all the problems. I didn't have to replace "defective memory" after all... Steve Link to comment Share on other sites More sharing options...
jquiros Posted November 6, 2003 Report Share Posted November 6, 2003 This is such BS and DANTZ mUSt know it. i KEEP getting elem.c-817 errors. i work for a large university. i've had this problem from the start. no, it was not my scsi card, no it is NOT memory, YES it IS a KNOWN issue with Retrospect that Dantz has acknowledged and hasn't dealt with since I puirchased the product. now the solution??? purchase an upgrade. come ON!!!!! what a crock! Link to comment Share on other sites More sharing options...
ericrow Posted November 10, 2003 Report Share Posted November 10, 2003 I have also been having elem-c 817 errors now and again. The latest one surfaced this weekend on a brand new set with new media on a windows client that has never caused errors before. I have seen these errors before and after upgrading to 5.1 on macs and pcs. This is excrementally annoying as retro just waits to be quit. How about a quick fix that just logs the error and keeps going? Is buying new RAM the only remedy? If so, what brands are recommended or are to be avoided? Is there any way to test if RAM is Retro-proof? My setup: OSX 10.2.7, 1 Ghz / 768 / AIT-2 FW Retrospect 5.1.175 Server. Link to comment Share on other sites More sharing options...
sparker Posted November 10, 2003 Report Share Posted November 10, 2003 I really, really hate to say it, but the assertion error is back for me too (after upgrading to 5.1). Without a lot of time to test it, it "appears" that it ONLY happens when backing up a Mac client (I backup my G3 as the host, a TiPB as a client, and a Dell Windoze Notebook as a client). I've only seen this error when working with the TiPB client. I'm not sure why, or what could be causing it, and since I typically have only a few hours at home each week to do my backups, I don't have a lot of time to spend testing. But to me, assertion errors indicate flawed (or "leaky") logic somewhere in the application. At least they're trapped relatively cleanly so the application doesn't tank entirely, but I have less and less faith in my backups every day. And faith is one thing REQUIRED in a backup scenario. Steve Link to comment Share on other sites More sharing options...
jquiros Posted November 13, 2003 Report Share Posted November 13, 2003 Welcome to the club of buying more things and still seeing the error. Do not trust their "go buy this and that" recommendations. It is a problem with the software, acknowledged by dantz. have you signed up to the list they have, dedicated to this problem (on which there have been NO emails since its inception). to sign up if you have faith (and i admit i still do), here's the info posted by AmyC on 5/29: =========== To be notified when there are updates on the elem.c-817 error, please subscribe to the following mailing list: http://list.dantz.com/mailman/listinfo/assert_elem817 =============== best of luck. i need it too. Link to comment Share on other sites More sharing options...
jquiros Posted December 16, 2003 Report Share Posted December 16, 2003 It really is frustrating, especially when you try calling in for assistance, too. then you're told your issue will be passed on to the customer support manager and that she will call you. the call never came. patience is at an end. dantz's support has proven to be terrible. here's one customer they've lost- including anywhere else he goes and anyone he recommends to. Link to comment Share on other sites More sharing options...
jquiros Posted December 16, 2003 Report Share Posted December 16, 2003 Now, the whole computer is just freezing. I swapped the whole computer and it happened again. The problem is that every time this happens- when Retrospect is using either of our two HP DLT-1 external drives and this happens. The drive has to be sent back to HP for repair!!!! HP's support, on the other hand, send a new drive by the next day. Dantz's support, in the meantime- really really just leaves you hanging. To this point, I can not express my frustration with Dantz enough. Link to comment Share on other sites More sharing options...
pnord Posted March 21, 2007 Report Share Posted March 21, 2007 This also crashes when I try to configure a client. Some still work. Most now crash Retrospect. Sorry guys... you had a fantastic product. I need to look for something else now. Paul Link to comment Share on other sites More sharing options...
pnord Posted March 21, 2007 Report Share Posted March 21, 2007 Correction: The client is Windows XP. The backup server is OS X 10.4.9. I'm backing up to a LaCie BiggerDisk Extreme. From the tremendous activity on this thread, I'd say that the problem is widespread. Since the problem isn't effecting all of my Windows clients just yet, perhaps it's a windows update that has just thrown a wrench into the works. I'll go check the update status on one of the functioning PC's. I'll let you know if it works after the update. Paul Quote: I'll agree with that. This problem just started showing up when backing up a particular OSX client. It crashes my backup server every time with "Assertion check at "elem.c-817". Can we get a fix soon? Please. Paul Link to comment Share on other sites More sharing options...
rhwalker Posted March 21, 2007 Report Share Posted March 21, 2007 Quote: From the tremendous activity on this thread, I'd say that the problem is widespread. ??? Yours is the first post to this thread in almost four years. There have been many updates and versions of Retrospect since then. If you are using a Retrospect version from 2003 on Mac OS 10.4.9, well, all bets are off. 10.4.9 requires Retrospect 6.1.x, which was not around four years ago. Russ Link to comment Share on other sites More sharing options...
CallMeDave Posted March 21, 2007 Report Share Posted March 21, 2007 Quote: From the tremendous activity on this thread, I'd say that the problem is widespread. LOL! The last post in this thread was in December of 2003; that's three and a half years ago! It's perfectly appropriate to resurrect an old thread if you're getting the same error. But at the very least, you need to provide _complete_ information about your configuration and experiences. Dave Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.