[Opensim-users] "Save OAR" stopping since r9700
BlueWall Slade
bluewall.slade at gmail.com
Sat May 30 07:36:06 UTC 2009
Rev 9603
http://opensimulator.org/mantis/view.php?id=3741
On Fri, May 29, 2009 at 3:45 PM, BlueWall Slade <bluewall.slade at gmail.com>wrote:
> Justin - I am setting http://opensimulator.org/mantis/view.php?id=3741up a
> local grid on 9395 and working up from there.
>
> Thanks!
> BlueWall
>
>
> On Fri, May 29, 2009 at 3:25 PM, Justin Clark-Casey <
> jjustincc at googlemail.com> wrote:
>
>> The most helpful thing is if you could go backwards from r9573 on Europa
>> until the problem goes away. And then slowly
>> inch forwards until it reappears, so we know which revision number the
>> issue is associated with.
>>
>> Charles Krinke wrote:
>> > I tried with r9573 this morning to do a "save oar <oarFileName> and here
>> > is my observation:
>> >
>> > First on ckTai which is essentially a blank region.
>> > Region (ckTai) # save oar ckTai0529
>> > 18:16:09 - [ARCHIVER]: Writing archive for region ckTai to ckTai0529
>> > 18:16:09 - [ARCHIVER]: 0 scene objects to serialize requiring save of 0
>> > assets
>> > 18:16:09 - [ARCHIVER]: AssetsRequest executed looking for 0 assets
>> > 18:16:09 - [ARCHIVER]: Creating archive file. This may take some time.
>> > 18:16:09 - [ARCHIVER]: Added control file to archive.
>> > 18:16:09 - [ARCHIVER]: Added region settings to archive.
>> > 18:16:10 - [ARCHIVER]: Added terrain information to archive.
>> > 18:16:10 - [ARCHIVER]: Added scene objects to archive.
>> > 18:16:10 - [ARCHIVER]: Wrote out OpenSimulator archive for ckTai
>> > Region (ckTai) # show version
>> > Version: OpenSimulator Server 0.6.4.9573 (interface version 3)
>> >
>> > Then on Europa which has pushing 1000 prims. What is interesting about
>> > Europa is that the first line (in bold) printed out immediately and then
>> > there is no response to console. The line below is 40 minutes later and
>> > there is still no console response.
>> >
>> >
>> > Region (Europa) # save oar Europa0529
>> > 18:18:03 - [ARCHIVER]: Writing archive for region Europa to Europa0529
>> > 18:31:43 - DIAGNOSTICS
>> >
>> > Time now is 5/29/2009 6:31:43 PM
>> > Server has been running since Tuesday, 5/19/2009 2:31:42 AM
>> > That is an elapsed time of 10.16:00:00.4956310
>> >
>> > ASSET STATISTICS
>> > Asset cache contains 1 assets
>> > Latest asset request time after cache miss: 0s
>> > Blocked client requests for missing textures: 0
>> > Asset service request failures: 0
>> >
>> > CONNECTION STATISTICS
>> > Abnormal client thread terminations: 0
>> >
>> > INVENTORY STATISTICS
>> > Initial inventory caching failures: 0
>> >
>> > FRAME STATISTICS
>> > Dilatn SimFPS PhyFPS AgntUp RootAg ChldAg Prims AtvPrm AtvScr
>> > ScrLPS
>> > 0.00 0 0.0 0.0 0 1 48 1
>> > 0 0
>> >
>> >
>> >
>> >
>> >
>> > ------------------------------------------------------------------------
>> > *From:* Justin Clark-Casey <jjustincc at googlemail.com>
>> > *To:* opensim-users at lists.berlios.de
>> > *Sent:* Friday, May 29, 2009 9:01:37 AM
>> > *Subject:* Re: [Opensim-users] "Save OAR" stopping since r9700
>> >
>> > The data that we (the developers) really need is the exact revision
>> > between r9561 to r9700 on which this started
>> > failing. I suggest a divide an conquer approach (see if the problem
>> > occurs on r9641, if not, then try r9670, etc.).
>> > Then that information should be put in a mantis (or an existing one, in
>> > which case I'll need to know the number).
>> >
>> > John Hopkin wrote:
>> > > Interesting - thanks (to both of you).
>> > >
>> > > I'm not what other information is relevant, but here's some:
>> > >
>> > > 1. For prims counts, they are two of the largest regions among the
>> > > seven, marked with asterisks here:
>> > >
>> > > 2889
>> > > 2717*
>> > > 858*
>> > > 526
>> > > 257
>> > > 195
>> > > 3
>> > >
>> > > 2. In total data size, they're the two of largest, as measured by the
>> > > size of the last good backup:
>> > >
>> > > 140M*
>> > > 31M
>> > > 23M*
>> > > 7M
>> > > 5M
>> > > 4M
>> > > 0.5M
>> > >
>> > > 3. Regarding terrain, as well as location on the grid, they're not
>> > > unusual; although they're adjacent to each other, all seven are
>> > > located contiguously and contain a mix of elevations.
>> > >
>> > > 4. They could well be the regions with the most scripts; I don't know
>> > > how I can find that out without a manual audit in-world.
>> > >
>> > > There are four other regions, not normally loaded, used occasionally
>> > > as sandboxes and storage areas. I don't yet know if they're affected
>> > > by the bug.
>> > >
>> > > Hope this gives some idea, and maybe helps. If you want to know
>> > > anything else, please ask.
>> > >
>> > > John
>> > >
>> > > BlueWall Slade wrote:
>> > >
>> > >> I have seen oar failures on Linux (Mono 2.4/64 Bit and Mono 2.5/32
>> Bit
>> > >> environments). And across several versions of OpenSim on several
>> > different
>> > >> regions. My experience is that any region, in these environments,
>> > with more
>> > >> than trivial content fails to complete the backup process. I am
>> > continuing
>> > >> to investigate. Any other information would be most helpful.
>> > >>
>> > >> Thanks,
>> > >> BlueWall
>> > >>
>> > >> On Thu, May 28, 2009 at 3:25 PM, dr scofield
>> > <drscofield at xyzzyxyzzy.net <mailto:drscofield at xyzzyxyzzy.net>>wrote:
>> > >>
>> > >>> John Hopkin wrote:
>> > >>>> Since I upgraded from r9561 to r9700, two out of seven regions
>> don't
>> > >>>> back up using "save oar". Even when they're the only region
>> running,
>> > >>>> those two have the same symptom: after a while, the archiver
>> appears
>> > >>>> to stop writing to the file, as if it's given up.
>> > >>>>
>> > >>>> Looking at the files with tar, they're fine for many MB, then the
>> > >>>> archive ends prematurely. There's nothing on the screen or in the
>> log
>> > >>>> to indicate a problem. This problem persists, and only on the
>> same
>> > >>>> two regions - other regions back up fine.
>> > >>>>
>> > >>>> Platform is Ubuntu Linux (Jaunty), Mono 2.2, MySQL 5.0; all seven
>> > >>>> regions run under one instance of OpenSim.exe.
>> > >>>>
>> > >>> my not actually be the issue, but mono 2.2 has a bad reputation.
>> 2.0.1
>> > >>> seems to be quite stable, 2.4 is also not too bad.
>> > >>>
>> > >>> dirk
>> > >>>> Any ideas? Thanks in advance.
>> > >>>>
>> > >>>
>> > >>> --
>> > >>> dr dirk husemann ---- math & computer science ---- ibm zurich
>> > research lab
>> > >>> RL: hud at zurich.ibm.com <mailto:hud at zurich.ibm.com> - +41 44 724
>> > 8573 - http://www.zurich.ibm.com/~hud/<http://www.zurich.ibm.com/%7Ehud/>
>> > <http://www.zurich.ibm.com/%7Ehud/><http://www.zurich.ibm.com/%7Ehud/>
>> > >>> SL: drscofield at xyzzyxyzzy.net <mailto:drscofield at xyzzyxyzzy.net>
>> > --------------------- http://xyzzyxyzzy.net/
>> > >>>
>> > >>> _______________________________________________
>> > >>> Opensim-users mailing list
>> > >>> Opensim-users at lists.berlios.de <mailto:
>> Opensim-users at lists.berlios.de>
>> > >>> https://lists.berlios.de/mailman/listinfo/opensim-users
>> > >>>
>> >
>> >
>> > --
>> > justincc
>> > Justin Clark-Casey
>> > http://justincc.wordpress.com
>> > _______________________________________________
>> > Opensim-users mailing list
>> > Opensim-users at lists.berlios.de <mailto:Opensim-users at lists.berlios.de>
>> > https://lists.berlios.de/mailman/listinfo/opensim-users
>> >
>> >
>> > ------------------------------------------------------------------------
>> >
>> > _______________________________________________
>> > Opensim-users mailing list
>> > Opensim-users at lists.berlios.de
>> > https://lists.berlios.de/mailman/listinfo/opensim-users
>>
>>
>> --
>> justincc
>> Justin Clark-Casey
>> http://justincc.wordpress.com
>> _______________________________________________
>> Opensim-users mailing list
>> Opensim-users at lists.berlios.de
>> https://lists.berlios.de/mailman/listinfo/opensim-users
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://opensimulator.org/pipermail/opensim-users/attachments/20090530/3d847eb1/attachment.html>
More information about the Opensim-users
mailing list