Home » Forums » MythDora » Usage

Hardware Upgrade Gone bad - Database confusion

  • user warning: Table './drupal_gding/cache_filter' is marked as crashed and should be repaired query: SELECT data, created, headers, expire, serialized FROM cache_filter WHERE cid = '1:4a9ac5493ea419ea76d3ea02719d81c8' in /var/www/mythdora/html/includes/cache.inc on line 26.
  • user warning: Table './drupal_gding/cache_filter' is marked as crashed and should be repaired query: UPDATE cache_filter SET data = '<p>Hello All,</p>\n<p>So I got brand new hardware that I am put on my network as a new Master Backend. Here is a quick overview, host Boron was my original backend/frontend using firewire to capture from my SA box. I also have host Beryllium as a frontend in my bedroom, which makes no further appearances from here on. All was well until I powered up host Neon, a rack mounted server that contains some new drives to be my primary network storage and encoding server.</p>\n<p>I installed 12.23 on Neon and migrated the database from Boron, first mistake. I have demoted Boron to a slave after the migration and after a few hours I have it recording the the NFS mount on Neon. I let things run for a while and noticed a few qwirks. </p>\n<p>Mythcommflag always fails with the following when ran for either backend, 192.168.1.10 is Neon and I have seen this use the loopback as well. I am confused as to why this doesn\'t use /storage/recordings/1649_20101115210632.mpg:<br />\n2011-03-08 19:47:08.580 RingBuf(<a href=\"myth://192.168.1.10:6543/1649_20101115210632.mpg\">myth://192.168.1.10:6543/1649_20101115210632.mpg</a>) Warning: Peek() requested 2048 bytes, but only returning -1</p>\n<p>My BE log is polluted with \"Should be local only query\" the log for today is at 40MB.<br />\n2011-03-08 08:41:09.096 Should be local only query: GetKeyframePositions - this logs during the entire length of a recording at about 29 per second<br />\n2011-03-08 03:45:06.469 Should be local only query: GetFramesWritten</p>\n<p>Mythweb never shows Thumbnail Icons, but this is not a big deal but I would like to fix it.</p>\n<p>The other thing is the ownership of my recordings is always root:root and some of my scripts need to be run as root which goes againt my good judgment.</p>\n<p>What it all comes down to, IMHO, is that I done hosed my stuff when I migrated the database. I am going to erase all my recordings and reinstall Neon for a clean start, but I would like to get some feed back on how to get back to a clean start without starting completely over. I would like for neon to be the master backend and record from a slave, which I don\'t see as a problem but if anyone has experience with this kind of setup I would love some feed back.</p>\n<p>If anyone know some tricks to get this working, by all means let me know so I can save some time.</p>\n<p>Thanks</p>\n', created = 1368997143, expire = 1369083543, headers = '', serialized = 0 WHERE cid = '1:4a9ac5493ea419ea76d3ea02719d81c8' in /var/www/mythdora/html/includes/cache.inc on line 109.
Submitted by beilber on March 8, 2011 - 6:08pm

Hello All,

So I got brand new hardware that I am put on my network as a new Master Backend. Here is a quick overview, host Boron was my original backend/frontend using firewire to capture from my SA box. I also have host Beryllium as a frontend in my bedroom, which makes no further appearances from here on. All was well until I powered up host Neon, a rack mounted server that contains some new drives to be my primary network storage and encoding server.

I installed 12.23 on Neon and migrated the database from Boron, first mistake. I have demoted Boron to a slave after the migration and after a few hours I have it recording the the NFS mount on Neon. I let things run for a while and noticed a few qwirks.

Mythcommflag always fails with the following when ran for either backend, 192.168.1.10 is Neon and I have seen this use the loopback as well. I am confused as to why this doesn't use /storage/recordings/1649_20101115210632.mpg:
2011-03-08 19:47:08.580 RingBuf(myth://192.168.1.10:6543/1649_20101115210632.mpg) Warning: Peek() requested 2048 bytes, but only returning -1

My BE log is polluted with "Should be local only query" the log for today is at 40MB.
2011-03-08 08:41:09.096 Should be local only query: GetKeyframePositions - this logs during the entire length of a recording at about 29 per second
2011-03-08 03:45:06.469 Should be local only query: GetFramesWritten

Mythweb never shows Thumbnail Icons, but this is not a big deal but I would like to fix it.

The other thing is the ownership of my recordings is always root:root and some of my scripts need to be run as root which goes againt my good judgment.

What it all comes down to, IMHO, is that I done hosed my stuff when I migrated the database. I am going to erase all my recordings and reinstall Neon for a clean start, but I would like to get some feed back on how to get back to a clean start without starting completely over. I would like for neon to be the master backend and record from a slave, which I don't see as a problem but if anyone has experience with this kind of setup I would love some feed back.

If anyone know some tricks to get this working, by all means let me know so I can save some time.

Thanks

‹ Alternate Web Browser to Firefox... very slow seeking ›