Jump to content


  • Content count

  • Joined

  • Last visited

Community Reputation

0 Neutral

About Mikee

  • Rank
    Retrospect Addict
  1. Mikee

    Version 7 and Exchange 2003

  2. My setup: Windows 2003 iSCSI initiator 1.06 Sony AIT2 Tape Drive connected to iSCSI switch (SanRAD V3000) Retrospect 7.0 demo My problem: Retrospect recognizes the tape drive. However, when it comes the time to write (erase tape) to the tape I get error -209 (Hardware Error). Any ideas? V7.0 lists iSCSI support, I wonder what exactly that means? Should iSCSI device be configured in certain way? thanks, Michael
  3. Mikee

    Mirror set gets errors

    "Target drives are removeable 250B EIDE drives. " Do you mean that you are using disk trays? Perhaps that's where the problem is. Try connecting the drives to the SATA controller via serial cable directly. Although not with SATA per se, I had problems with ATA trays in the past. Mikee
  4. High CPU usage also comes from Proactive backups. They ( 4 of them, 83 clients each) rutinely run my cpu at 100% while checking for source/polling ). Unfortunatelly there's nothing you can do but hope that the next version will be programmed better. As far as compression goes - if you have dual processor and Win 2000 and up you could enable ntfs compression instead of using Retrospect's for better performance. I run some tests in the past and found that standard 4k cluster size is the most optimal for file backups. ( I, however, no longer use file backups as they were getting corrupted whenever HD run out of disk space ) Mikee
  5. Mikee

    Mirror set gets errors

    I had this problem with Promise FastTrack Tx2000 where mirror would go out of sync (no data corruption though). Promise tech support claimed that this is normal at high speeds (RAID10 configuration). I switched to 3ware 7500 and did not have any problems since. Mikee
  6. Create advanced selector that selects folder name (ie. folder name exactly matches My Program). If folder name is not unique than you'll need to in use path selector (ie. folder path ends with \Program Files\My Program). Unfortunatelly current Retrospect cannot backup just one folder. You'll be forced to either backup the selected folder and no security+attributes information (Source Snapshots OFF) or the folder and all security+attributes of the whole partition (Source Snapshots ON) ( can take hours if you have a lot of files on the drive). There's no workaround. This is by design. Mikee
  7. I run between 3-8 execution units. In my experience number of execution units has little to do with CPU usage. It is cumulative backup speed of all execution unit what maxes up the processor. My backup PC (1GB ram, 2.8 P4, software compression) runs into 100% CPU utilization at about 930MB/min backup rate. Mikee
  8. Mikee

    Dual CPU with Mutli-Server

    Hi. 12% of 1Gb is 14.6MB/s, that is about 878MB/minute. Since network bandwidth is not a problem -It appears that something within Retrosoect is limiting the throughput. thanks, Mikee
  9. Mikee

    retrospect server troughput on win xp?

    I do see the same thing: about 12% (120Mb) max. utilization of 1Gb. My server is W2k3 w/2.4GHz P4 and Intel GB NIC. I think the bottle-neck is not the network but the software itself. My (wild) guess is that compression engine is limiting the throughput. I did not try it myself (as is useless to me as I don't have enough storage), but perhaps you can try disabling software compression to find out if throughput increases. Btw, according to my tests 4k cluster size is faster than 16k for Retrospect Disk backups. (by about 8% in my case) good luck, Mikee
  10. Mikee

    Dual CPU with Mutli-Server

    hm...are you sure that 1GHz is enough? I'm running up to three jobs at the time and backup to IDE RAID10 and Firewire drives using 1Gbit NIC and my CPU is at 99% when network utilization reaches 12% of 1Gb. My setup is 2.8GHz P4 with 1GB of ram on i865 chipset. Btw, Does anyone post higher throughput than 120Mbit/s while backing up? Mikee
  11. Windows Task Manager, at performance tab under Commit Charge (K) keeps track of memory usage. Total is your allocated memory (ram+swap). Limit is the maximum available (ram+swap). Peak will tell you the highest amount of ram+swap ever allocated since boot time. First , make sure that swap has the disk space to grow to it's maximum value. Next watch the Peak value. If's close to the Limit you'll need to add more ram or increase swap size...if not...look somewhere else for a fix... hope that helps, Mikee
  12. The second log example is correct. I made a mistake in my previous post saying that the correct value is 32.1GB. I ment to type 36.1GB. Mikee
  13. I'm experiencing similar kind of behavior when transferring separate clients out of Disk Set onto a Tape Set. There's no bandwidth issues here as the Disk Set is on local RAID10 disk array. Even the random writes are above 6MB/s threshold. The tape stops and goes. Retrospect is doing Please Wait-Updating destination-Updating Catalog File-Copying routine. And, the transfer speed hovers at around 28.0MB/minute... ...as compared to 400MB/min if whole set is being transfered at once. I guess I could create intermediate Disk Set with just those clients, but why? Dantz has to really optimize this...please. Mikee
  14. You're right, it seems like the size of the backup is being misreported (32.1 GB is the correct value) so that compression ratio doubles and in turn the speed doubles as well climing to impressive, but incorrect, >2000MB/minute sometimes I usually don't pay attention to compression ratio. I trust that if (1) backup had no errors, (2) verification found nothing wrong- the backup was successful. Shouldn't that be the case? thanks, Mikee
  15. Example: (same server, same Storage Group) - 2/1/2004 3:24:54 AM: Copying First Storage Group on K3 2/1/2004 4:58:51 AM: Comparing First Storage Group on K3 2/1/2004 5:42:22 AM: Execution completed successfully Completed: 3 files, 72.1 GB, with 66% compression Performance: 1088.6 MB/minute (794.2 copy, 1730.9 compare) Duration: 02:17:28 (00:01:57 idle/loading/preparing) - 2/3/2004 3:31:18 AM: Copying First Storage Group on K3 2/3/2004 5:08:35 AM: Comparing First Storage Group on K3 2/3/2004 5:53:21 AM: Execution completed successfully Completed: 3 files, 36.1 GB, with 33% compression Performance: 527.4 MB/minute (384.2 copy, 840.5 compare) Duration: 02:22:02 (00:02:06 idle/loading/preparing) I assume it's just a reporting bug. Isn't it? thanks, Mikee