Search:

Type: Posts; User: Mike McConnell

Search: Search took 0.01 seconds.

  1. Replies
    2
    Views
    11,653

    I stand corrected on item #1. It appears that...

    I stand corrected on item #1. It appears that amstatus has nothing to do with the renaming. When amdump completes it renames the amdump file to amdump.1 and increments the suffix on all amdump.#...
  2. Replies
    2
    Views
    11,653

    Misc items

    1) After a run completes, if you run amstatus <config> it will rename the amdump and log files by adding the timestamp. Once this happens you cannot run amreport because it cannot find the files.
    ...
  3. Replies
    16
    Views
    15,155

    Success! The complete 1.6T run spanning 5...

    Success!

    The complete 1.6T run spanning 5 tapes finished normally. I haven't had a chance to check the results and try a restore on a couple of files yet but I'd be surprised if there was a...
  4. Replies
    16
    Views
    15,155

    Got some very interesting results that makes me...

    Got some very interesting results that makes me believe we're definitely in the hunt but more tuning is required.

    I did some investigation into the suggestions Dustin made and found what looked to...
  5. Replies
    16
    Views
    15,155

    Paul, I looked at the manufacturers web page...

    Paul,

    I looked at the manufacturers web page and they suggested 256 and other posts I read in the forum suggested improved performance. As I mentioned, 32K buffers actually was a bit better...
  6. Replies
    16
    Views
    15,155

    Thanks Dustin. I'll do some more investigation. ...

    Thanks Dustin. I'll do some more investigation.

    Attached in the taper debug log with some of the middle snipped out to reduce its size. One reoccurring entry at the failure point strikes me as...
  7. Replies
    16
    Views
    15,155

    That all makes sense. Thanks. The dump failed...

    That all makes sense. Thanks.

    The dump failed with the same Kernel error message :confused:

    Mike
  8. Replies
    16
    Views
    15,155

    Update: I printed and studied the logs and saw...

    Update:

    I printed and studied the logs and saw something that I had seen before but shrugged off since the other config for small dumps was working.

    I had previously fought a battle to get the...
  9. Replies
    16
    Views
    15,155

    There is some relationship but I'm not sure what...

    There is some relationship but I'm not sure what to make of it. When I used 20G split size It wrote 8G chunks, when I used 4G is wrote 5G chunks. Since I've only run both single DLE and the 6 part...
  10. Replies
    16
    Views
    15,155

    Yes I see the same error. It made a smooth...

    Yes I see the same error.

    It made a smooth transition from the first part, picked up and continued on tape MCNET-001 and continued with the next block sequence number from where it left off until...
  11. Replies
    16
    Views
    15,155

    tried breaking DLE up

    So in desperation I tried breaking the DLE into smaller pieces and as I suspected, it didn't help. From amstatus:

    MCS1:/export/1-B 0 238g wait for dumping
    MCS1:/export/C-D 0 301g...
  12. Replies
    16
    Views
    15,155

    tape spanning issue with large DLE

    Sorry for the long post in advance.

    I'm having problems with tape spanning of very large DLEs.

    Background Info:
    client machine: 3.8G Extreme HT x86_64, Centos 5.2 (64bit), 4G Ram, 2.5T SATA II...
  13. I'm sure the block size was an issue as it...

    I'm sure the block size was an issue as it certainly helped me but now I have another problem. When I run full backups, which are significantly larger than my holdingdisk capacity, I get extremely...
  14. I found the solution after googling this to near...

    I found the solution after googling this to near death! :)

    All that is necessary was to add one entry in /etc/modprobe.conf file as follows.
    options st buffer_kbs=256

    The answer came from here...
  15. As is usually the case, I manged to formulate a...

    As is usually the case, I manged to formulate a reasonable google search and found a couple of hints about using LILO to change the boot options. More digging and I found that st.buffer_kps=256 would...
  16. Forgot to mention OS is CentOS 5.2. Thanks, ...

    Forgot to mention OS is CentOS 5.2.

    Thanks,

    Mike
  17. LTO3 performance improvements & blocksize

    I complied and have 2.6.0p2 up and running and I'm trying to improve the performance. The tape is moderately slow and from past experience I know it's capable of much more.

    I searched the forums,...
  18. Replies
    2
    Views
    2,744

    creating a client rpm (or package)

    I've been working with 2.6.0p2 and built from source. I have the server running and have been working on fine tuning but the clients I've been using we're downloaded from the CentOs (5.2) repository...
  19. Replies
    3
    Views
    3,077

    taperflush also has the inequality backwards too....

    taperflush also has the inequality backwards too. Sorry I missed it in the the first post.

    Mike
  20. Replies
    3
    Views
    3,077

    flush-threshold (s) confusion

    I'm confused. The man amanda.conf documentation clearly states:

    " flush-threshold-dumped int
    Default: 0. Amanda will not begin writing data to a new volume until the
    ...
  21. Replies
    2
    Views
    3,446

    Thanks

    Thanks for the confirmation Dustin. I knew that was the likely answer I was just hoping for a silver bullet to make it go away.

    I'll fetch the current rev and go from there.

    Thanks again.
    ...
  22. Replies
    2
    Views
    3,446

    TCP Port Ranges

    I have been working on Amanda for about two weeks and have been partially sucessful. I can backup and use amrecover sucessfully on the amandahost machine (MCS1) so the basics are up and running. I...
Results 1 to 22 of 22