AF:
NF:0
PS:10
SRH:1
SFN:
DSR:
MID:<20080725162540.358781e3@ripper.onstor.net>
CFG:
PT:0
S:andy.sharp@onstor.com
RQ:
SSV:onstor-exch02.onstor.net
NSV:
SSH:
R:<lauren.quilici@onstor.com>,<dl-CougarCore@onstor.com>
MAID:1
X-Sylpheed-Privacy-System:
X-Sylpheed-Sign:0
SCF:#mh/Mailbox/sent
RMID:#imap/andys@onstor.net@onstor-exch02.onstor.net/INBOX	0	BB375AF679D4A34E9CA8DFA650E2B04E0B0C6FE9@onstor-exch02.onstor.net
X-Sylpheed-End-Special-Headers: 1
Date: Fri, 25 Jul 2008 16:26:17 -0700
From: Andrew Sharp <andy.sharp@onstor.com>
To: "Lauren Quilici" <lauren.quilici@onstor.com>
Cc: "dl-Cougar Core Team" <dl-CougarCore@onstor.com>
Subject: Re: Minutes from 7/22 Cougar Core Team Meeting
Message-ID: <20080725162617.393b4caf@ripper.onstor.net>
In-Reply-To: <BB375AF679D4A34E9CA8DFA650E2B04E0B0C6FE9@onstor-exch02.onstor.net>
References: <BB375AF679D4A34E9CA8DFA650E2B04E0B0C6FE9@onstor-exch02.onstor.net>
Organization: Onstor
X-Mailer: Sylpheed-Claws 2.6.0 (GTK+ 2.8.20; x86_64-pc-linux-gnu)
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit

Just one update from these notes, due to testing carried out by Jon, we
determined that killing just one FP core wasn't enough, we will be
killing two, and even then our stout little cougar cub will be faster
than we really want.  If you had told me I would one day say that
phrase I would have called you a silly git.

Cheers,

a

On Fri, 25 Jul 2008 15:49:53 -0700 "Lauren Quilici"
<lauren.quilici@onstor.com> wrote:

> All,
> 
>  
> 
> If you missed Tuesday's meeting, then you can see what was presented
> by using the following link:
> 
> <\\mightydog\Program Management\Cougar\Weekly Team
> Meetings\07-2008\07-22-2008
> <file:///\\mightydog\Program%20Management\Cougar\Weekly%20Team%20Meeting
> s\07-2008\07-22-2008> >
> 
>  
> 
> Attendees:  Brian M., Dennis, Eugene, Jobi, John R., Jonathan, Larry,
> Lauren, Mike B., Paul H., Paul N., Rich, Vikas
> 
>  
> 
> HW Status
> 
> 1.	Paul N. reported that the pilot build of 35 systems was
> started at Venture on 7/21 (71 total motherboards).  The first 5
> completed systems are expected by 7/29. 
> 2.	The 1u chassis is still being worked.  It has passed EMI and
> safety testing.  The target manufacturing release date is 7/25.  We
> can GA anytime after that. 
> 3.	Jonathan was concerned with the loose cable connection
> found on the unit at Shopzilla.  Paul N. reported that about 80% of
> the boards have slight left/right movement but it's within the
> tolerance of the chassis. 
> 4.	Paul N. reported that we had 2 suppliers for DIM connectors.
> One was disqualified because the extractor mechanism is soft and the
> connector tends to get loosened in transit.  We won't see this problem
> going forward. 
> 5.	Paul also reported that the compact flash ejector has been a
> problem since Bobcat.  We need to retrain Venture to take care of
> this. 6.	Larry reported that the "sc1 not pingable" issue
> could be a HW problem on Rev 4 boards (transmit side of sc1 port).
> He will discuss this with Brian S. 
> 
>  
> 
> Dev Status
> 
> 1.	Jonathan reported that there are around 20 GA MF defects. 
> 2.	The dump problem is the big issue left.  We know there is no
> small workaround for this.  Jonathan said Max needs 2-3 days (starting
> 7/21) to create a solution.  This will only affect NDMP.  We will not
> have to retest NDMP or RMC, just a dataset that's known to fail. 
> 3.	Jonathan also reported that we dropped the ball on the
> 6520.  He has Andy working on a Linux kernel change and Warren
> working on a PROM change.  There will be a single version of PROM for
> all platforms that keys off the model number.  The code will turn off
> 1 FTP core and cut down memory in the 6520.  Vikas stated he thinks
> this is too risky to change without testing because it could change
> the timing.  Paul H. stated we can include the code in systems we
> sell, but not activate it until it's been tested.  Paul N. reported
> this is a separate discussion and we won't ship without testing. 
> 4.	Larry reported that there are 12 - 15 defect fixes for
> sub32. Vikas requested a submittal today and another on Friday. 
> 5.	Paul N. reported we will load the RC on Mightydog on 7/28. 
> 
>  
> 
> QA Status
> 
> 1.	Vikas reported that QA testing on sub31 is going well and a
> lot of progress has been made. 
> 2.	The restore issue has been resolved. 
> 3.	QA regression is roughly 70-75% complete. 
> 4.	For Cougar, major features under testing now are Mirror,
> DMIP, and Backup/Restore.  QA got the VirusScan CDs from John and
> plan to make progress on this and Security Explorer this week.  For
> 3.3, the only testing left is for Security Explorer, cluster testing,
> and cluster upgrade testing. 
> 5.	SS has been up for 4 days and the defect find rate has
> declined. CS was fully up to load with 22 clients but the main server
> was rebooted so tests need to be restarted.  The configuration is
> complete.  So far no new defects have been found with the larger
> load. 6.	QA is down to 64 total defects. 
> 7.	Vikas reported that there is still a lot left for QA to do
> and other activities (LSI, Beta installation) are consuming
> resources.  HCL is helping, but he's still concerned. 
> 
>  
> 
> Mightydog Update
> 
> 1.	Vikas reported that only one new issue was found on
> Mightydog last week.  John confirmed this and reported that the EEK,
> performance and corruption issues are still under investigation. 
> 2.	John reported that he has not added the additional storage
> to Mightydog yet.  Bill is taking a look to see if he thinks there
> will be a problem.  Paul H. reported that Jonathan gave Amit a test
> to run.  The result will determine if we add the storage or not.
> John will talk this over with Amit. 
> 3.	Jonathan reported that the performance problem seems to be
> highly correlated to snapshot creation and is expected behavior. 
> 
>  
> 
> Qual Team Status
> 
> 1.	Brian M. reported that Fay is done with Nexsan (V100)
> testing, LSI testing, and all of Bob's requests.  He hasn't started
> the Fujitsu 4-node testing yet because Dave has run into some path
> optimization issues in his testing with Fujitsu's new FW. 
> 2.	Paul H. told Brian we need the SPEC FS numbers before GA and
> Vikas agreed that will meet his needs as well.  John is using SPEC FS
> for LSI array-based snapshot work.  Paul said this is almost done and
> they will talk separately to allocate resources. 
> 3.	John reported that Shopzilla equipment is out of the boxes
> and ready to be racked, but the rail kits weren't sent back.  Brian
> said he'd ask Shopzilla for these. 
> 4.	Brian reported they are still working with IBM and hope to
> finish that testing in a couple weeks. 
> 5.	Dave will fix the LSI Chrystal failover control issues
> after he finishes with Fujitsu. 
> 
>  
> 
> Beta Update
> 
> 1.	Paul N. reported that we have shipped 11 Cougar units so
> far. All units are loaded with sub28, except 3 at SGI and the unit at
> Shopzilla (until Raj gets there tomorrow), which have sub26.  All
> other units are currently on ship hold pending investigation of a
> variety of install issues.  We need to tell ONStor Japan not to
> install any units at SGI until we solve our install problems here.
> We want the complete unit back from ONStor GmbH for failure analysis. 
> 2.	Paul reported that a Beta restart program was initiated
> over the weekend.  We're now attempting to make every install follow
> the same process.  A Dev Engineer will be assigned to each Beta unit
> and Paul H. will distribute a report from them on Friday of each
> week.  The first 5 units from the pilot build will be prepared for
> shipment using current processes.  They will then be installed at
> ONStor by CS and all issues encountered will be catalogued.
> Hopefully by next week we'll be on top of lots of these issues. 
> 3.	Vikas reported that Beta defects will be assigned out of the
> triage meeting.  He will send out a list before the meeting on
> Wednesday.  Issues found at particular sites can be found on Vikas'
> slides. 
> 4.	Jonathan reported that there seems to be a consistent time
> skew on the Beta machines.  Paul N. stated the time is programmed at
> ICT. Vikas found the time to be one hour behind at one install, but he
> couldn't reproduce the problem in the lab.  Paul believes this is a
> DST issue and will look into it. 
> 5.	Rich pointed out that the NTP issues we're having with Beta
> machines could be due to this clock issue. 
> 6.	Paul N. reported that both nodes are currently up and
> running and clustered at ZDF.  There are issues with sc0 (not
> pingable, packet loss).  John reported the install team is losing the
> heartbeat.  This is a possible HW issue. 
> 7.	Paul N. reported that all units shipped so far are Beta
> units and should be treated as such.  He believes only 1 Beta
> contract has been signed.  We will initiate an RMA program when it's
> time to get the units back, unless someone is interested in buying. 
> 
>  
> 
> Documentation
> 
> 1.	Dennis reported he needs to know what defects we'll publish
> as known problems for 4.0.  His documentation needs to be sent out for
> review next week so he needs the public tab updated by Monday.  Paul
> H. said Andy, Vikas and Rich should decide.  Vikas agreed to run a
> query through the list of known defects to find candidates. 
> 2.	Dennis already has the queries for 3.3 set up and he will
> work on that list. 
> 
>  
> 
> Please feel free to "Reply All" if I've missed anything important!
> 
>  
> 
>  
> 
> Lauren Quilici
> Project Coordinator
> 
> ONStor, Inc.
> office: 408.376.3137
> 
> lauren.quilici@onstor.com <mailto:lauren.quilici@onstor.com> 
> http://www.onstor.com <http://www.onstor.com> 
> 
>  
> 
