Well I found a fix that works, currently testing. J-D
On Tue, Nov 3, 2009 at 2:15 PM, Jean-Daniel Cryans <[email protected]> wrote: > Ananth, > > If you can, please try the patch included in HBASE-1781 > (https://issues.apache.org/jira/browse/HBASE-1781) > > Thx, > > J-D > > On Thu, Oct 29, 2009 at 6:41 AM, Ananth T. Sarathy > <[email protected]> wrote: >> No idea. >> Ananth T Sarathy >> >> >> On Thu, Oct 29, 2009 at 5:59 AM, Andrew Purtell <[email protected]> wrote: >> >>> The FindBugs static code checker flagged WildcardColumnTracker.checkColumn >>> also, null dereference possible on some code paths. I checked in a fix which >>> did not hurt this issue but would not help either. I added null tests and >>> flow control to prevent the issue seen by the checker but did not change (if >>> I was succesful in my intent) any semantics. That code should be >>> refactored/rewritten for easier comprehension and cleaner flow control. >>> >>> - Andy >>> >>> >>> >>> >>> ________________________________ >>> From: stack <[email protected]> >>> To: [email protected] >>> Sent: Thu, October 29, 2009 1:35:27 PM >>> Subject: Re: stackoverflow error >>> >>> Whoever wrote that checkColumn code was obviously trying to be a >>> smarty-pants but missed a vital check. Any of you fellas have an idea what >>> the columns are that are causing this infinite recursion? Otherwise will >>> poke around tomorrow to see if can create it artificially. >>> >>> Thanks, >>> St.Ack >>> >>> On Wed, Oct 28, 2009 at 6:38 PM, Gary Helmling <[email protected]> >>> wrote: >>> >>> > I just updated to 0.20.1 and think that I'm seeing this as well. >>> > >>> > Here's a sample of the client log: http://pastebin.com/m549fbcfa >>> > and a sample of the regionservers: http://pastebin.com/m5a8e89c3 >>> > >>> > Best I can tell, it seems to be data related in that it's reproducible >>> with >>> > gets on specific rows. Other rows in the same region work normally. >>> > >>> > >>> > >>> > On Wed, Oct 28, 2009 at 9:25 PM, Jean-Daniel Cryans <[email protected] >>> > >wrote: >>> > >>> > > BTW this is https://issues.apache.org/jira/browse/HBASE-1781 >>> > > >>> > > J-D >>> > > >>> > > On Wed, Oct 28, 2009 at 6:24 PM, Ananth T. Sarathy >>> > > <[email protected]> wrote: >>> > > > It's not advancing. I am not sure how you tell which region... >>> > > > here is some of the output from the shell >>> > > > http://pastebin.org/49134 >>> > > > >>> > > > >>> > > > Where can I look for that info >>> > > > Ananth T Sarathy >>> > > > >>> > > > >>> > > > On Wed, Oct 28, 2009 at 6:24 PM, stack <[email protected]> wrote: >>> > > > >>> > > >> Same region over and over? Then it ain't advancing? Or does it >>> vary? >>> > > >> St.Ack >>> > > >> >>> > > >> On Wed, Oct 28, 2009 at 3:09 PM, Ananth T. Sarathy < >>> > > >> [email protected]> wrote: >>> > > >> >>> > > >> > ok... I when I try to do i straight countin the shell >>> > > >> > I get >>> > > >> > >>> > > >> > >>> > > >> > 09/10/28 18:08:56 DEBUG client.HTable$ClientScanner: Advancing >>> > > internal >>> > > >> > scanner to startKey at >>> > > >> > >>> > > >> > >>> > > >> >>> > > >>> > >>> 'FileZilla\x203.2.33.2.3http:\x2F\x2Ffilehippo.com\x2Fdownload_filezilla\x2F5408\x2 >>> > > >> > >>> > > >> > which has been repeating for a while... >>> > > >> > Ananth T Sarathy >>> > > >> > >>> > > >> > >>> > > >> > On Wed, Oct 28, 2009 at 5:08 PM, stack <[email protected]> wrote: >>> > > >> > >>> > > >> > > You using a filter of some kind? What kinda scan is it? Are >>> you >>> > > on >>> > > >> > > latest >>> > > >> > > hbase? Easy way to check healthyness is scan of the table. Use >>> > > count >>> > > >> in >>> > > >> > > shell or rowcounter mr job. If you are just doing straight >>> scan, >>> > > then >>> > > >> > > something is up. Looks at the 0.20.1 bug fixes to see if any >>> > > address a >>> > > >> > > stackoverflow issue. Otherwise, try to figure why its happening >>> > for >>> > > >> you >>> > > >> > > and >>> > > >> > > not for the rest of us. >>> > > >> > > St.Ack >>> > > >> > > >>> > > >> > > On Wed, Oct 28, 2009 at 1:51 PM, Ananth T. Sarathy < >>> > > >> > > [email protected]> wrote: >>> > > >> > > >>> > > >> > > > I am getting a stackoverflow error while I am iterating >>> through >>> > a >>> > > >> table >>> > > >> > > > scan >>> > > >> > > > >>> > > >> > > > http://pastebin.com/mc940416 >>> > > >> > > > >>> > > >> > > > is there any maintiance type task I should run to make my >>> > > >> regionservers >>> > > >> > > are >>> > > >> > > > all copacetic? >>> > > >> > > > >>> > > >> > > > Ananth T Sarathy >>> > > >> > > > >>> > > >> > > >>> > > >> > >>> > > >> >>> > > > >>> > > >>> > >>> >>> >>> >>> >>> >> >
