cblundell Posted June 20, 2007 Report Share Posted June 20, 2007 Hi, I'm running a small backup scenario with 2 iMacs & 1 Macbook. We have a dedicated server (a G4).. The setup has been running fine for a year or so... the 2 imacs were auto backed up, and the macbook manually. All backups are file backups, which alternate their saving to 'Backup set A Hard Drive' & 'Backup set B Hard Drive' (2 100gb drives in the G4) My client has recently moved offices, and so we've started backing up again... except the macbook is now in the office enough to be included in an auto backup so i added it.. My problem is that the macbook is occasionaly creating the error 206. This has only happened on the macbook. the error looks like: "Trouble reading: “Hobart Backup Set B” (28666335), error 206 (drive reported a failure: dirty heads, bad media, etc.)." There are multiple of these errors.. they have been during auto & manual backups... To try and alleviate this, i de-installed the client on the macbook and reinstalled a fresh copy with the new version. This might have done the trick (time will tell) But could someone help as to why i could be getting this error ? The drives are physically ok (no read/write errors, no smart status warning etc) All i can find is that error 206 relates to Tape drives & dvd/cd Burners NOT file backups.... Can Anyone help with this one ? Thanks Charlie. Quote Link to comment Share on other sites More sharing options...
twickland Posted June 20, 2007 Report Share Posted June 20, 2007 Quote: The drives are physically ok (no read/write errors, no smart status warning etc) All i can find is that error 206 relates to Tape drives & dvd/cd Burners NOT file backups.... In my experience, the 206 error is always generated by trouble with the destination volume or drive. As you note, it's typically reported by the drive mechanism itself when there is trouble reading or writing to the media. However, I can confirm from experience that hard drive I/O errors can also generate this message--sometimes thousands of 'em--both for file and removable disk backup sets. Retrospect can really stress hardware, and drive mechanical or communication problems can be revealed in Retrospect long before they appear under other, less demanding means of accessing the drive. Are you getting the error only for Hobart Backup Set B? Quote Link to comment Share on other sites More sharing options...
CallMeDave Posted June 20, 2007 Report Share Posted June 20, 2007 Quote: the error looks like: Where in the backup process is the error occuring? Is it during the Verify stage? Dave Quote Link to comment Share on other sites More sharing options...
cblundell Posted July 16, 2007 Author Report Share Posted July 16, 2007 Hi, Apologies for the slow response... Yes it does look like it was during verifying. I've now turned verify off - and of course that stops the error - but probably not any underlying issues. What is interesting is that this error was being reported on files on the HardDrive we call HardDrive B. I have had other (possibly unrelated) errors on this hd, namely - the backup has 'got out of sync' with the database, and so i've had to 'rebulid' it. Could the 206 error and the out-of-sync error be indicative of HardDrive Bad bits ? or could it be something else ? The other HArd drive we use (HD A) has had no problems. (for the record, i have deleted the backup file, and made a new one when i was originally having problems to ensure the file itself wasn't corrupt) Thanks for your help. Quote Link to comment Share on other sites More sharing options...
CallMeDave Posted July 23, 2007 Report Share Posted July 23, 2007 Quote: What is interesting is that this error was being reported on files on the HardDrive we call HardDrive B. I have had other (possibly unrelated) errors on this hd, namely - the backup has 'got out of sync' with the database, and so i've had to 'rebulid' it. Could the 206 error and the out-of-sync error be indicative of HardDrive Bad bits ? Yes. Since both are internal drives in the G4, how are they connected? Do they share a bus? Is one a slave and another a master? You might want to get a nice PCI host adapter for the drive, and assure each is a master on it's own bus/channel. Dave Quote Link to comment Share on other sites More sharing options...
cblundell Posted July 23, 2007 Author Report Share Posted July 23, 2007 Hi Dave, Are you suggesting IT IS Definitely a/some bad blocks... and that i should consider this the answer to fixing the problem ? or are we just assuming that it is most probably bad blocks ? (I appreciate that this is a hard call to make, but would be good to get an idea as to how certain we are that this is the issue.) The next part of my question (& i'm not sure of the setup at the moment) is how would having a separate bus' for each hard drive (apart from a small possible speed increase) affect the backups ? It obviously wont affect data reliability... nor help in reducing hard drive failure... if you could explain further that'd be great.... Thanks for your help in nailing this down.... for the record i'm gonna replace with 2 fresh 500gig drives, and will hopefully map out any bad blocks of the failing drive, and use as extra user storage. Cheers, Charlie. Quote Link to comment Share on other sites More sharing options...
mhegge Posted July 30, 2007 Report Share Posted July 30, 2007 Similar issues with latest Retrospect Workgroup backing up to external Lacie-brand FW800 drives. I have worked with Lacie, replaced or repaired drives and still come up with the "dirty hards error:. What I end up doing and keep re-running the script until it finally completes a backup. Sometimes it will fail immediately and sometimes it will back up 100GB or more before erroring out. Always the same error. I beleive its software related in Retrospect, purely because there has not been a significant upgrade to the software in ages. I have another setup on a G5 XServer backing up an Xraid to te same type of Lacie externals with no issues. Problem Server: OSX Server 10.4.9 (soon to be 10.4.10) PowerMac G4 Dual 867, 1.5GB RAM, 80GB HD Sonnet FW800 PCI card Lacie Extreme BIG externals (1.2 TB and 2TB) connected via FW800 Quote Link to comment Share on other sites More sharing options...
CallMeDave Posted July 30, 2007 Report Share Posted July 30, 2007 Quote: I beleive its software related in Retrospect, purely because there has not been a significant upgrade to the software in ages. Oh bollocks. Gut feelings are a great place to start. But emprical testing is what's necessary to prove or dis-prove a theory. I don't see you describe if you are using File Backup Sets stored on the external drive, or if you are using the external drive as a device for Removable Backup Sets. But either way, when the software reports a hardware error, a good first testing step would be to swap out for some _other_ hardware! Especially when the hardware you're using is known to have these sorts of issues, with users from all over the world reporting problems with power supplies and cooling. Quote Link to comment Share on other sites More sharing options...
mhegge Posted July 30, 2007 Report Share Posted July 30, 2007 I have two servers in two separate locations. Location one is a G5 Xserver with a XRAID running Retrospect Workgroup locally and backing up to two 1.2GB Lacie Big D2 Bigger Disk Extreme via FW800. Daily backups have been flawless for over a year with no "dirty heads errors" In my second location, I have a PowerMac G4 DUAL 867, 1.5GB RAM, Dual 80GB drives and one 320GB drive with two partitions, running OS X Server 10.4.9, internal Sonnet FW800 card and SIIG Internal eSATA card, external 5-Bay RAID NITRO AV eSATA. Retrospect Workgroup running locally. eSATA RAID and a 1.2GB Lacie external house the main shared data. One script backs up data from one partition that has a share to the other partition. This backup is consistently successful. A second script reads from the eSATA RAiD share, an external Lacie 1.2GB drive share and backs up to the external Lacie 2TB. This is the backup I have had trouble with. I have had the backup drive replaced twice, reformatted a couple of times, had all drives including the eSATA RAID tested and rebuilt using DiskWarrior and/or Techtool Protege. I am not thoroughly convinced it is not hardware, specifically the Lacie drives, heat or throughput issues. I just want some progress in software by EMC regarding Mac OS X ADAN maybe some insight from someone having similar issues with similar hardware. Quote Link to comment Share on other sites More sharing options...
CallMeDave Posted July 30, 2007 Report Share Posted July 30, 2007 Quote: I have two servers in two separate locations. Location one is a G5 Xserver with a XRAID running Retrospect Workgroup locally and backing up to two 1.2GB Lacie Big D2 Bigger Disk Extreme via FW800. Daily backups have been flawless for over a year with no "dirty heads errors" This seems to indicate pretty strongly that even though there has not been a Retrospect update in a long while, the existing software build works reliably with good hardware. That also bolsters the theory that there is something wrong with the volume where you are saving your File Backup Set. FW800 host adapter, FireWire bridge-board, Lacie RAID controller, Lacie power supply, etc. EMC can't solve hardware problems with software updates. Dave Quote Link to comment Share on other sites More sharing options...
fuzzydisco Posted August 23, 2007 Report Share Posted August 23, 2007 I was about to post on a problem very similar to this... I am encountering two variations of the error 206. The first setup includes a La Cie 1.2 TB drive connected as a network disk via Airport Extreme backing up automatically once a week to a La Cie 2 TB drive connceted via FW800 to my G5 tower. Just a normal backup script, nothing too fancy. It worked perfectly for several months until this week's backup, now it stops early into the copying phase with the 206 error. Nothing has changed, and the drive passes all hardware tests! The second scenario involves trying to back up a PB G4 in firewire target disk mode to DL DVDs through a mac mini--basically mounting the PB's drive on the mini, then attempting to run a backup to removable media. I have also attempted this same scenario using my G5 tower in place of the mac mini, both get the same result, a 206 error early on in the script process. Since it is doubtful that these are all hardware failures, any more ideas what could be causing this? Quote Link to comment Share on other sites More sharing options...
rhwalker Posted August 23, 2007 Report Share Posted August 23, 2007 It all depends on what you call "hardware failures". Unless you use special software, it's difficult to get SMART status on any but directly-connected drives (i.e., those without "bridge" chips between one type of interface - e.g., firewire - and another - ATA, etc.). In the case of firewire drives, it could be one of several issues: (1) firmware in the "bridge" chip used in the drive enclosure to translate between the firewire bus and whatever external bus the drive has - solution would be to update the firmware, if possible, or replace the enclosure. (2) power supply issues for the case - LaCie is known for marginal power supplies. It's possible that your first case is caused by a power supply that has developed a lot of noise on one of its voltages. If it's a drive with an external brick supply, you might try swapping out that power brick. (3) firmware in the drives themselves (4) firewire driver issues (5) etc., etc. Retrospect has no control over any of these. Now, I have no personal knowledge of your setup, and I haven't looked at your logs for errors from the various drivers. But Retrospect doesn't talk directly to firewire devices - it just makes standard syscalls and reports the errors. Russ Quote Link to comment Share on other sites More sharing options...
CallMeDave Posted August 23, 2007 Report Share Posted August 23, 2007 Quote: Since it is doubtful that these are all hardware failures I haven't seen enough information here to doubt hardware as being the fault. First, what exactly are the messages you are getting? Are they "can't read" or are they "can't write" messages? > It worked perfectly for several months until this week's backup OK, so the software/hardware worked before, but now it does not. So what might have changed? It's unlikely that anything in the Retrospect application changed between then as now, in such a way as to cause an error of the type generally associated with bad hardware. You could try with fresh preferences, or even reinstall the program. But if that doesn't fix it, you should focus on non-software related causes. > both get the same result, a 206 error early on in the script process Both are writing to optical media, yet you omit any information about the make/model/driver of the optical drive. You also omit any information about the version of Retrospect or the RDU it's loading. All in all, a "me too" post without enough additional information doesn't do much to solve anything. Dave Quote Link to comment Share on other sites More sharing options...
peterthorpe Posted August 28, 2007 Report Share Posted August 28, 2007 We had the same 206 errors and initially found an odd fix. File Backup failed after a few GB (in Copying mode), with the 206 error. The problem was solved, temporarily, by changing the backup to a Removable Disk backup (using partitions on the same backup disk). We had no 206 errors for several months, but suddenly they are back (again the backup fails after a few GB in Copying mode). Our system: Mac: Dual 2 GHz PowerPC G5 running OS X 10.4.10 Backup disk: Various external hard drives (mostly Lacie) - 500GB up to a 2TB (partioned); we have tried switching power bricks or drives with no clear effect. Connected via Firewire 400, 800 or USB (no difference). Retrospect:6.1.126 all drivers/firmware are up to date. Quote Link to comment Share on other sites More sharing options...
CallMeDave Posted August 28, 2007 Report Share Posted August 28, 2007 Quote: We had the same 206 errors Are they "can't read" or are they "can't write" errors? Quote Link to comment Share on other sites More sharing options...
peterthorpe Posted August 29, 2007 Report Share Posted August 29, 2007 Seems to be a write problem (see error below). Trouble writing: “1-August07” (2025328128), error 206 (drive reported a failure: dirty heads, bad media, etc.). Quote Link to comment Share on other sites More sharing options...
mharter Posted August 31, 2007 Report Share Posted August 31, 2007 I got another set of 500GB drives for my 2 backup servers and they all have dirty heads according to Retrospect. We have been using the WD MyBook drives for a while now and have had no problems until now. I have tested the drives using various utilities and they check out just fine. I could order more drives and hope that they work, but that would be up to chance This is a nightmare for us, and we'll be looking for a product to replace Retrospect as soon as possible. Reliability is so important when it comes to backing up data. Matt Quote Link to comment Share on other sites More sharing options...
CallMeDave Posted August 31, 2007 Report Share Posted August 31, 2007 Quote: I got another set of 500GB drives for my 2 backup servers and they all have dirty heads according to Retrospect. This post comes out of the blue, with no background or information about hardware or software used. So it's impossible to know what's happening, other then to suggest that most users don't see this sort of error when working with well performing hardware. When you write "We have been using the WD MyBook drives for a while now and have had no problems until now" does that mean you were using that hardware with Retrospect without problem? Or just that you were using that hardware in general? And what about the backup servers? What model of Macintosh are they? I personally stay away from WD drives. I don't think that they are as good as Seagate or even Maxtor. But as usual, more information in the post yields more helpful answers from the Forum community. Dave Quote Link to comment Share on other sites More sharing options...
mharter Posted August 31, 2007 Report Share Posted August 31, 2007 OK, I'll be specific, and I am sorry about the lack of background information. Retrospect server version 6.1.126 running on mirror door Mac with OSX 10.4.6, 768MB RAM, USB2 card model 1PC-U2OPC5N-04 WD5000I032-001 Works fine WD5000C032 does not work, 602 dirty heads after about 3 hours Retrospect server version 6.1.126 running on G5 1800 single processor Mac using OSX 10..4.6, 512 RAM, built in USB2 WD5000I032-001 Works fine WD5000C032 does not work, 602 dirty heads after about 3 hours Thanks! Matt Quote Link to comment Share on other sites More sharing options...
CallMeDave Posted August 31, 2007 Report Share Posted August 31, 2007 So the same device (WD5000C032) running on two independent machines, reports the same hardware based error in the same manner. Judging from the facts presented, I'd say that there is an issue with the model WD5000C032. I couldn't find any information on this specific model on the WD website except for one reference to a replacement power supply. And since power supplies for external drive enclosures often are the source of problems (since they are often poor quality with limited current available) perhaps this is a good place to start. Quote Link to comment Share on other sites More sharing options...
rhwalker Posted August 31, 2007 Report Share Posted August 31, 2007 Quote: Retrospect server version 6.1.126 running on mirror door Mac with OSX 10.4.6, 768MB RAM, USB2 card model 1PC-U2OPC5N-04 WD5000I032-001 Works fine WD5000C032 does not work, 602 dirty heads after about 3 hours Retrospect server version 6.1.126 running on G5 1800 single processor Mac using OSX 10..4.6, 512 RAM, built in USB2 WD5000I032-001 Works fine WD5000C032 does not work, 602 dirty heads after about 3 hours might be instructive to swap the power supplies on the good and bad drives, see if the errors stay with the drives or with the power supplies. Could also be a thermal problem from this problem description. Remember, Retrospect is at a higher level than the drives, and talks to the drives through OS routines and just reports errors, and Retrospect will beat on the drives and demand continuous-duty operation, which the drives might not be designed to handle. I think you are barking up the wrong tree if you think this is a Retrospect issue. Sounds like a hardware issue with the particular drives. Russ Quote Link to comment Share on other sites More sharing options...
mharter Posted September 4, 2007 Report Share Posted September 4, 2007 Thanks for the suggestions. The power supplies are not likely to be the issue, I swapped them with the ones that are proven to work but doing so did not make a difference. It is possible however that the newer WD5000C032 drives use more power and over work any of the power supplies. I have since bought new drives and cases, and have had them running during the week end. They succeeded in backing up our data. The drives I purchased are Seagate ST3500641AS-RK drives installed into Eagle/Consus W-series (no name) drive cases. I was hesitant about using Seagate drives because I've had huge failure rates by them in the past when I bought a lot of them for use as internal drives. I hope they've gotten better. As for using no name drive cases, what is the alternative, Lacie? I haven't found them to be any better than anything else. We hammer on our Western Digital Mybook drives as transfer drives, they are great for copying hundreds of gigs of video files, why doesn't Retrospect like them? Is there a list of Retrospect certified drives available? I'm not trying to be critical, I'm just trying to save others hundreds of dollars and hours of labor. Matt Quote Link to comment Share on other sites More sharing options...
peterthorpe Posted September 27, 2007 Report Share Posted September 27, 2007 still trying to track down the 206 error! I am working to eliminate various possible issues. We have one backup system that works fine and I notice that when I check "about retrospect" from the retrospect menu and then select "Status" then I see "Configuration: 128 K" (amoungst other things) However, on our other system (which is throwing up the 206 errors) the same Status indicates "Configuration: 79 K". In fact this seems to change with time usually in the 70-80K range. I have no idea what "Configuration" means here - is it indicative of memory usage? I thought OS X provided automatic memory allocation? Can I manually set this value? While I doubt that this could be causing the 206 errors, it is one of the few differences between the two backup machines we have running and I would like to check that it is not the source of our problems. Any help would be appreciated. Thanks ps. as before our systems are: Mac: Dual 2 GHz PowerPC G5 running OS X 10.4.10 Backup disk: Various external hard drives (mostly Lacie) - 500GB up to a 2TB (partioned); we have tried switching power bricks or drives with no clear effect. Connected via Firewire 400, 800 or USB (no difference). Retrospect:6.1.126 Device access version: 1.0.107 Driver update version: 6.1.10.100 (I know there are updates, but this version seems to work on one of our systems) Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.