I had this. Replacing the "continue" on line 224 with "return" does the trick.

Thanks!
Well, is this a real bug? Will this be changed in the next update? Or is the "continue" there on purpose?
I will probably wait for a official fix rather than changing this on my own.

The poller was split up yesterday. This is a bug from that.

There are a few issues, and Mike will more than likely have them fixed quickly.

If you do change it yourself, when you get a merge conflict, simply deleting this file then pulling again will be enough to fix it.

Okey thank you for the advice! I won't doubt that this will be fixed as fast as anything happens here on Friendica :D

Yeah that was fixed about 12 (?) hours ago.

It wasn't - it's doing the same thing later on in the code (your fix was at line 110, this is 224)

Should have all of them now.

I see that at free-haven (in the error log), and the site will not load at all. I just pulled again, and rebooted the server, even, and the site will not load. Everything else on the same server, including statusnet, etc., is all working fine, but trying to get to free-haven.org, it just hangs and hangs and hangs...I don't even get an error page or time-out, or anything, just "connecting...waiting for free-haven.org..." forever.

I've got lots of this:
28-Apr-2012 23:35:23] PHP Warning: Invalid argument supplied for foreach() in /var/www/freehvn/friendica/include/plugin.php on line 71
[28-Apr-2012 23:35:23] PHP Warning: Invalid argument supplied for foreach() in /var/www/freehvn/friendica/include/plugin.php on line 71
[28-Apr-2012 23:35:23] PHP Warning: Invalid argument supplied for foreach() in /var/www/freehvn/friendica/include/plugin.php on line 71
[28-Apr-2012 23:36:40] PHP Warning: Invalid argument supplied for foreach() in /var/www/freehvn/friendica/include/poller.php on line 153
[28-Apr-2012 23:37:48] PHP Warning: Invalid argument supplied for foreach() in /var/www/freehvn/friendica/addon/facebook/facebook.php on line 1211
[28-Apr-2012 23:37:49] PHP Warning: Invalid argument supplied for foreach() in /var/www/freehvn/friendica/include/queue.php on line 89
[28-Apr-2012 23:40:06] PHP Fatal error: Cannot break/continue 1 level in /var/www/freehvn/friendica/include/onepoll.php on line 224
[28-Apr-2012 23:41:19] PHP Fatal error: Cannot break/continue 1 l... show more

line 224 of include/onepoll.php is just
return;

Line 224 was continue; until a few hours ago. The cannot break/continue error is fixed.

The other error message is a different problem. Since all the other errors are about plugins, have you tried moving the addon directory outside of your Friendica install? If that works, you can add them back one at a time until you find the offending one.


(ETA, to be honest, what I do at times like this is rm -r and reclone the git)

All the other errors shown there were from before the pull I just did. The site loaded, now, but it took over 5 minutes to get the login page. I logged in, and it's been 2 or 3 minutes waiting for the site to load again.

Tony - something wrong with the 'addon' table. That's not one we've messed with recently. Check to see if it crashed.

Everything is loading very quickly except anything that involves viewing posts (home, network, community pages).

Those are taking 5 to 10 minutes, if they load at all.

And then I see no posts. All content gone.
The directory is intact, my contacts are intact, but no content in the public stream loads at all.

Take a deep breath, and look at your item table.

check if you have a column 'archive' in your contact table. If not, perform the sql in update_1138

What is my item table, and how do I "perform the sql in update_1138"?

In your database, check the table called "Item".

You've got the same symptoms as when mine corrupted.

Though Mike is telling you check a known issue, so check that first.

I logged into mysql (mysql -u root -p), and then
mysql> use freehvn;
Reading table information for completion of table and column names
You can turn off this feature to get a quicker startup with -A


hanging...hanging...I think there's something hosed with my db.
(clearly, "freehvn" is the name of the db for my friendica installation).

Now, I was over at Quinnipiac U. on my laptop a few hours ago and on free-haven, and all was fine.
I ssh-ed to the server with my phone and did a pull (couldn't get ssh through Quinni-U's guest wifi, or something), when I read about the FB connector polling timing isue thing, and, I swear I was still using the site after that, briefly. Then I put the laptop away and went to a quiet corner and read my Anne McCaffrey book for a couple of hours, and came home to find these issues.
I pulled a while ago, now, after reading this thread.

Tried to restart mysql, and got this error:
ERROR: The partition with /var/lib/mysql is too full! ... failed

There's definitely something very wrong going on.
I was only using about 10gb of the 20gb storage on my server last I checked, like a week ago, and now
df -h says I'm using 100% of the disk, too.

This is crazy.

Take a look at /var/log/mysql...

/var often has logfiles. Might be one of these (that isn't being rotated) filling it up.

../friendica/log/friendica.log (my friendica log) is 6225360249 bytes.

/var/log/mysql is completely empty.

10873993 lines in friendica.log

Yeah, I don't use logs 90% of the time.

I've set it up, I just need to touch the appropriate file when I need them, and rm when I've got the bits I need. I also turn them on for 24 hours when there's an update to Facebook, diaspora.php or something like that.

The rest of the time, you can't find anything in logs that big anyway.

@Darth Baldwin Time to dump the log, methinks.

Ok, erase your friendica log file and turn Friendica logging off. Restart server, and various services should load. Then, at least your drive won't be full and you can get saner error from mysql (or hopefully, none).

hdd was too full. This was largely the product of logs, both friendica.log, and /var/log/statusnet/statusnet.log
The SN log as 6gb, and the friendica.log was immense.
I cleared these out, and about 2/3 of the older apache logs, and went from 20gb back down to 6.5 gb

Then I was able to restart mysql again, however, I did this error:
ERROR 144 (HY000) at line 1: Table './freehvn/photo' is marked as crashed and last (automatic?) repair failed

I'm getting content, now, but all my photos are gone.
I turned of the FB connector, momentarily.

Could someone tell me what "perform the sql in update_1138" entails, or at least point me to where I can find out?

update_1138 is the most recent update in update.php.

function update_1138() {
q("alter table contact add archive tinyint(1) not null default '0' after hidden, add index (archive)");
}

This isn't likely to be the problem now that we know what the problem is. Incidentally, I have to purge my Friendica logfile about every 3 days or it gets so large that I can't open it.

But if you look at update.php you'll see a bunch of numbered functions that each perform some update operation - usually on the database, to keep up with code changes.

If an update fails (and you'll soon start getting emails if this happens) you may need to look at the SQL query for that particular update and possibly manually execute that SQL in mysql to see why it borked.

By the way, did you try

mysql> repair photo;

??

mike wrote:

now that we know what the problem is

I'm not entirely clear, yet, and what the problem is. Or, an overfull hdd is definitely a problem, but, what's the problem with my photos table? How can I repair it? I lost all photos, and don't seem to be able to upload one, now, to replace my userpic, even.

mysql> alter table contact add archive tinyint(1) not null default '0' after hidden, add index (archive);
ERROR 1060 (42S21): Duplicate column name 'archive'

This concerns me more:
mysql> describe photo;
ERROR 144 (HY000): Table './freehvn/photo' is marked as crashed and last (automatic?) repair failed

tried to repair
mysql> repair photo;
ERROR 1064 (42000): You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near 'photo' at line 1

Ah...supposed to be "repair table photo;"

My bad, I don't use it very often...

repair TABLE photo;

That seems to have helped. I was able to replace my profile photo, and all my photos are back!

I still wonder what caused the photo table to be corrupted.
Could that be a symptom of the overfull hdd?

Yes. Tried to write a block, the write failed, tried to report it, but the write failed, tried to gracefully recover but the write failed. At some point it becomes hard to do anything but throw your hands up and die.

I've gone into /etc/logrotate.d and set all the big stuff (apache, mysql, fail2ban) to
rotate 4
maxage 10
I [i]believe[/ir] (if I've understood correctly), this will only keep up to 4 logs for any given app, and won't keep any log older than 10 days.
Hopefully this will prevent logs from piling up and filling the hdd.