Jump to content


  • Posts

  • Joined

  • Last visited

Everything posted by PGalati

  1. Is it possible to tell Retrospect to only poll a specific IP or a range of IP addresses to backup? I would like to try only backing up the computer when it is connected to the built-in ethernet port and avoid backing up over wifi to reduce traffic there. Or can I tell the client plist file to hard code a specific IP address. Thanks.
  2. Started noticing a rash of emails notices this week indicating that Retrospect had launched, but I never shut it down. So last night I get a notice saying the backup can't run now because it can't find the backup set. Is it toast? [*] soccRecv: recv failed, error 60 Assert occurred on 4/4/18 at 2:43:26 PM + Retrospect version Launched at 4/4/18 2:43:27 PM Assert occurred on 4/4/18 at 3:28:15 PM + Retrospect version Launched at 4/4/18 3:28:16 PM + Normal backup using File Server Backup (Manual) at 4/4/18, 10:00:00 PM (Activity Thread 1) Can't access Backup Set 091917, error -1,101 (file/directory not found) 4/4/18 10:00:12 PM: Execution incomplete Model Name: Mac Pro Model Identifier: MacPro4,1 Processor Name: Quad-Core Intel Xeon Processor Speed: 2.93 GHz Number of Processors: 2 Total Number of Cores: 8 L2 Cache (per Core): 256 KB L3 Cache (per Processor): 8 MB Memory: 16 GB
  3. Thanks for the response and sorry for the delay. I don't think I remember seeing the notification of a reply to my post. Anyway... The catalog file at this point is reaching 54GB in size. The media set size is 6TB containing 1.4 million files across 15 members. Each time I start over and essentially create a new media set, everything works very well initially. As time goes by the beach ball starts to reveal itself as the catalog file, media set size, and file count start to increase. My guess is it is the sheer fact of writing a 56GB catalog file after each workstation backup. Even thought the catalog is on an SSD, it still takes quite awhile. I just timed it and I get the spinning beach ball for just about 10 minutes before the app becomes responsive again, only to go unresponsive moments later as it connects to the next workstation to look for changes to back up. My past backups currently show 4066 since Sept 19, 2017. Each workstation is backed up M-F, once a day.
  4. I am experiencing a situation where the Retrospect app on my laptop tends to go unresponsive for sometimes up to a minute before becoming available again for a few moments. It seems each time it writes to the 36GB catalog file on a 16 Intel Zeon core Mac Pro with SSD, the remote app on my MacBookPro gets the spinning beach ball for a considerable amount of time. I can continue working in other apps without delay but simply have to wait for Retrospect to return. The engine does continue to function because when the app returns to life, it may be involved in another workstation backup already. I am not sure why the catalog file is so large. I am backing up roughly 30 workstations and 1 Mac OS X server once a day to a single catalog to avoid duplication between server and workstation. This catalog file was created on Sept. 19th 2017. Obviously with an empty catalog file, the response times for the app were quick. As time passes and the catalog takes on some data, the not responding starts to get longer and longer, probably due to the growing size of the catalog. Currently the catalog does get compressed and no grooming is configured. All data is saved on hard drives, no tape or optical. Is there something that could be done to reduce the size of the catalog file? Do most admins use separate catalogs for each client? Are there articles that describe best case scenerios based on computer environment config? Thanks for your time.
  5. I tried performing a couple searches to find info about my situation but maybe my search terms aren't providing the info I am looking for. I have a Mac OS X server with 2 nic cards installed, one for local gigabit ethernet access and the other is for 3 drobo iSCSI based storage devices. I put the local ethernet as priority 1 and the drobo nic as priority 2. I recently rebooted the server and when the server came back up, the Retrospect client attached itself to the IP address of the iSCSI network and not the local ethernet interface. The local is 192.168.1.x and the iSCSI is 10.0.100.x. The network obviously has no knowledge of the 10.0 network so the Retrospect is now giving a 519 error not able to find the agent. I think the only way I can currently fix this is to reboot with the iSCSI network unplugged so the retro client could attach to the 192 network. I did have in preferences "allow clients to turn off retrospect software" turned off, so I cannot do that manually until I get the client reconnected to the retro server. Is there a way to force the client to connect to a particular nic regardless of a ready state. Both ethernet cables were plugged in during the boot process and boot are configured as manual static addresses. Any suggestions would be greatly appreciated.
  6. Even after removing the .session files from each of the members folders, the result is still the same. I click rebuild, get an empty catalog file, and no disk access ever takes place. At this point since Retrospect is the only thing corrupted and not any production servers, I am simply starting over at this point. Will implement the duplicating of the catalog file during an idle time to have a duplicate in case of future corruption.
  7. OK, so tech support suggested that I rebuild the catalog file. I copied all of the members of the media set to a single hard drive for hopefully an easier rebuld. If I understood correctly from tech support, the folder structure should look like this: 6TB -Retrospect --03-01-16 ---1-03-01-16 ---2-03-01-16 up to ---8-03-01-16 When I choose rebuild and select disk, it asks to add members. I point to the Retrospect folder and it replies with the name of the script not encrypted. It then asks me where I want to save the rebuilt catalog to. Regardless of where I save it to, I get a message indicating that this is an existing media set, ok to replace? My only choice forward is to choose Rebuild. When I do that, I get a catalog file called 01-16 and it is empty and no disk activity ever takes place. Log file looks like this: + Executing Rebuild at 7/22/16, 3:51:49 PM (Activity Thread 2) To Backup Set 01-16... Using thorough rebuild 7/22/16 3:51:51 PM: Finished scanning backup set data files 7/22/16 3:51:51 PM: Execution completed successfully I have gone ahead and created a new media set in the meantime so I can continue backing up new data but I would like to resume this media set if I can. Any suggestions as to what my next step might be? Thanks, Paul
  8. I was adding a driver to support an 3/4 expresscard SATA port multiplier on our MacBook Pro Retrospect Server. After installing the driver I was asked to reboot. I installed this driver at 6PM. Now when I open Retrospect I see in the log there was this statement at 4:17PM. Now my media set is visible but it contains zero of the 8 or so members that were there yesterday. The catalog is 3.13GB in size. 7/19/16 4:17:45 PM: Compressing catalog file 7/19/16 4:17:52 PM: Compressing catalog file necoSoccTimer: connection failed (idle 301 seconds) UAllowSystemSleepAgain: IOPMAssertionRelease failed, err = -536,870,208 Exit at 7/19/16 6:02:55 PM E-mail notification failed: error -597 (mail server not found) NetRespTop::nereCreateClientFromNode: Importing client 'Quickbooks-PC' (11.0.0 (252)) NetRespTop::nereCreateClientFromNode: Importing client 'VOIP' (11.0.0 (252)) NetRespTop::nereCreateClientFromNode: Importing client 'Laptop' (11.0.0 (252)) NetRespTop::nereCreateClientFromNode: Importing client 'MS-Project-PC' (11.0.0 (252)) Each time I restart the computer now , Retrospect logs + Retrospect version Launched at 7/20/16 10:10:46 AM UAllowSystemSleepAgain: IOPMAssertionRelease failed, err = -536,870,208 Exit at 7/20/16 10:25:30 AM E-mail notification failed: error -597 (mail server not found) + Retrospect version Launched at 7/20/16 10:25:58 AM I get the launch email but not the exit email. Is this a retrospect issue or a Mac OS X 10.6.8 issue? I am on hold for tech support but thought I would post this here as well. Thanks for any knowledge you can provide. Paul
  9. I am not sure where or if suggestions are received but I have one that would be efficient for admins keeping the clients up to date. While having the Retrospect app open, it provides a notification to the admin "I just finished successfully backing up the client "admin's Mac" but I notice that the client is out of date. Would you like me to update the client from to Or better yet, have Retrospect provide a preference (On or Off) that would auto update the client based on the version queried on the client computer at the time of backup and the version located in the Retrospect's Client updater folder on the Retrospect server. With laptops coming and going off the network, it is sometimes a task to keep up with finding the window of opportunity to update the client before the client disappears again. Anyway, just a thought. Thanks. Paul
  10. Even though my preferences say the past backups are sets to 500, my past backups indicator shows 2000 and growing. Is that an expected behavior because retrospect is also periodically becoming unresponsive for around 30 seconds after every successful backup. this happens both on the retro server or on a remote retro server.
  11. Retro v12 did not even attempt to make a connection to the mail server. There is no evidence of Retro making a connection in the mail server's log file. When I clicked "Send Test E-mail" it immediately brought up the error message, and no connection attempt exists in the mail server log. Only after disabling authentication did it begin to work again. UPDATE: In my particalur case, it appears that it is not the authentication that causes it to fail, it is the SSL. I have auth turned on but SSL turned off and it does work. Having SSL on immediately creates a "No server found at that address (-597) when I click Send Test E-mail. I did have Auth and SSL on in v11 and it worked as expected, now SSL in v12 does not like my username and password. For the record, the outgoing mail server is listed as dotted IP and not as mail.domain.com, but again, the dotted IP and SSL on worked as expected in Retro v11. I enabled debugging up to 7 in the secret preferences based on the KB article but it does not provide any more info about SSL failing when attempting to send a Test E-mail. It simply provides the same error message listed in preferences. Thanks, Paul
  12. Thanks for the reply. Looking at the mail logs from Monday when Retro was still running 11.5, it does appear that Retro was properly authenticating with the mail server. 11:30:38.348 2 SMTPI-247735([]) authenticated as postmaster@domain.com 11:30:38.348 4 SMTPI-247735([]) rsp: 235 postmaster@domain.com relaying authenticated Now after installing v12, I did not see a single reference in the mail logs referring to a connection coming in from Retro UNTIL I turned off smtp auth. 11:10:08.775 4 SMTPI-260476([]) cmd: HELO ip99.domain.com 11:10:08.775 4 SMTPI-260476([]) rsp: 250 domain.com we trust you ip99.domain.com 11:10:08.776 4 SMTPI-260476([]) cmd: MAIL FROM:<Retrospect@domain.com> 11:10:08.776 4 SMTPI-260476([]) rsp: 250 Retrospect@domain.com sender accepted 11:10:08.777 4 SMTPI-260476([]) cmd: RCPT TO:<pgalati@domain.com> As shown, no authentication was negotiated but accepted since ip99 is whitelisted. I am gettng a "No Server Found at that Address (-597)" error, and it never initiates an outbound connection. Again, turning off smtp auth fixes it.
  13. Not sure if others have experienced this but my email notifications started failing with a -597 error after updating to v12. Once I removed the checkbox to require smtp authentication, the test email was sent successfully. Reenabling authentication caused it to fail again. Since I manage the mail server as well I am able to work around this issue, but wanted to know if others may have seen this. Nothing changed in the email preference between 11.5 and 12 other than it stopped working as configured. The user name is just the account name not including the domain and the password.
  14. In the past couple weeks, a strange thing has happened to Retrospect. I created a new script with new external hard drives to backup our Tiger Server with xServe Raid. The first time it runs the backup, it basically does a full backup, which is normal. The next night it performed a normal backup which is incremental, so it backs up what changed, maybe 2GB worth. Last night, it decided that it needed to back up 160GB of data. What happened? What did someone logged into the file server do to make Retrospect think that it needed to back up that much data? The reason I created a new script is because the prior working script that was a year old with no such occurences, started doing this as well. None of the sets are flagged as "missing". Retrospect Backup 6.1.126 Mac OS X 10.3.9 Any suggestions are greatly appreciated.
  • Create New...