AND...slave was promoted to a master that database did not come up
right and I cannot connect to it.

As I prefer not to lose all data again I will currently try to leave
that at this situation (while the new slave starts and might not be
able to complete initialization due to this situation) and wait a
while if someone can respond. Don't really know where to take it from
here: kill mysql on the new master, just restart it regularly (might
fail), reboot, shutdown?

On Jan 26, 9:48 am, afishler <[email protected]> wrote:
> After losing the database on Saturday completely I started at new.
>
> Currently I am facing almost EXACT the same issue as 
> herehttp://groups.google.com/group/scalr-discuss/browse_thread/thread/b1c...
>
> only that my db totally hangs.
>
> To me it seems almost clear that there is an issue of the data bundle
> with database activity.
>
> After restarting I left the db idle or mostly doing read actions. When
> I regained db population the problem surfaced again.
>
> I say again that this is a real issue. As far as I am concerned I
> cannot rely on this solution.
>
> When is the new release coming out?
>
> Arie.
>
> On Jan 24, 6:48 pm, Arie Fishler <[email protected]> wrote:
>
> > also...I just noticed, and this is the worst case now...the new master
> > extracted a snapshot successfully. All db files seems to be in place....db
> > even shows me the list of tables and still any select action on the tables
> > returns with an error that THE TABLE DOES NOT EXIST.
>
> > I dont really know what is this status of the database and whether it is
> > totally lost at this point. slave behaves exactly the same.
>
> > On Sat, Jan 24, 2009 at 6:29 PM, Arie Fishler <[email protected]> wrote:
> > > not exactly. the slave was 100% stuck in "not being the master". the 
> > > second
> > > slave that started did not succeed in starting (several started and 
> > > failed).
> > > I tried to shutdown mysql on the slave before executing the script ->
> > > failed
> > > tried executing the script -> failed (it also tried to stop mysql)
> > > only then rebooted.
> > > mysql did not start saying it executed slave2master on a non slave.
> > > Probably the reboot did initiate the process to become a master but the 
> > > fact
> > > I ran it manually confused it.
> > > I was left with no master up....then terminated all instances and sclar
> > > automatically started instances right.
> > > Naturally there was no db during this period.
>
> > > I will keep following.
>
> > > On Sat, Jan 24, 2009 at 6:14 PM, Nickolas Toursky <[email protected]>wrote:
>
> > >> Slave should promote itself is a master automatically. If it has
> > >> failed, the only way to do it - to execute /usr/local/aws/bin/mysql-
> > >> slave2master.sh script in the shell.
> > >> As I can see, you have rebooted the master instance before it was
> > >> initialized - this has broken this instance and did not allow slave to
> > >> initialize properly.
>
> > >> On Jan 24, 10:19 am, afishler <[email protected]> wrote:
> > >> > Hi,
>
> > >> > Currently after losing the master again, the slave is not promoted to
> > >> > be a master. How do I do it manually?
>
> > >> > New instance that is starting does not seem to survive and keeps
> > >> > starting over.
>
> > >> > This is real frustrating. I really don't think this is an issue of
> > >> > occasional disk issues on AWS. Master dying happens around twice a
> > >> > day....too frequent to be considered a "normal" problem
>
> > >> > Here is the log of the new slave failing to start
> > >> > 24-01-2009 01:46:56     ERROR   i-f9800290/instance-up.sh
> > >> /usr/local/
> > >> > aws/bin/mysql-init.sh failed. Exiting.
> > >> > 24-01-2009 01:46:55     INFO    i-f9800290/mysql-init.sh
> > >>  Traceback (most
> > >> > recent call last):
> > >> > File "/usr/bin/s3cmd", line 415, in
> > >> > error("S3 error: " + str(e))
> > >> > File "/usr/lib/python2.5/site-packages/S3/S3.py", line 41, in __str__
> > >> > retval += (": %s" % self.info["Code"])
> > >> > KeyError: 'Code'. Retrying.
> > >> > 24-01-2009 01:46:55     ERROR   i-f9800290/mysql-init.sh        Could
> > >> not fetch
> > >> > MySQL data snapshot using index s3://farm-1173-918348349691/farm-mysql/
> > >> > mysql-snapshot.tar.
> > >> > 24-01-2009 01:46:55     ERROR   i-f9800290/mysql-init.sh        Failed
> > >> to fetch
> > >> > 's3://farm-1173-918348349691/farm-mysql/mysql-snapshot.tar' to '/mnt/
> > >> > mysql-misc/tmp.TySmbs2490/mysql-snapshot.tar' for 4 tries.
> > >> > 24-01-2009 01:46:24     INFO    i-f9800290/mysql-init.sh
> > >>  Traceback (most
> > >> > recent call last):
> > >> > File "/usr/bin/s3cmd", line 415, in
> > >> > error("S3 error: " + str(e))
> > >> > File "/usr/lib/python2.5/site-packages/S3/S3.py", line 41, in __str__
> > >> > retval += (": %s" % self.info["Code"])
> > >> > KeyError: 'Code'. Retrying.
> > >> > 24-01-2009 01:46:04     INFO    i-f9800290/mysql-init.sh
> > >>  Traceback (most
> > >> > recent call last):
> > >> > File "/usr/bin/s3cmd", line 415, in
> > >> > error("S3 error: " + str(e))
> > >> > File "/usr/lib/python2.5/site-packages/S3/S3.py", line 41, in __str__
> > >> > retval += (": %s" % self.info["Code"])
> > >> > KeyError: 'Code'. Retrying.
> > >> > 24-01-2009 01:45:53     INFO    i-f9800290/mysql-init.sh
> > >>  Traceback (most
> > >> > recent call last):
> > >> > File "/usr/bin/s3cmd", line 415, in
> > >> > error("S3 error: " + str(e))
> > >> > File "/usr/lib/python2.5/site-packages/S3/S3.py", line 41, in __str__
> > >> > retval += (": %s" % self.info["Code"])
> > >> > KeyError: 'Code'. Retrying.
> > >> > 24-01-2009 01:45:53     INFO    i-f9800290/mysql-init.sh        Trying
> > >> to fetch
> > >> > previous MySQL snapshot from s3://farm-1173-918348349691/farm-mysql/
> > >> > mysql-snapshot.tar (10527569920 bytes).
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"scalr-discuss" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to 
[email protected]
For more options, visit this group at 
http://groups.google.com/group/scalr-discuss?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to