SEQanswers

Go Back   SEQanswers > Sequencing Technologies/Companies > Illumina/Solexa



Similar Threads
Thread Thread Starter Forum Replies Last Post
multiple cartridges in a single MiSeq run? wingtec General 1 08-10-2012 01:21 PM
BWA fail to open file wisosonic Bioinformatics 12 06-06-2012 01:46 AM
How to Demultiplex a Nextera paired-end MiSeq run allo Illumina/Solexa 6 02-27-2012 08:10 AM
cuffmerge fail papori Bioinformatics 0 07-31-2011 04:01 PM
Cuffdiff outputs mostly FAIL agc Bioinformatics 0 01-11-2011 05:45 AM

Reply
 
Thread Tools
Old 03-06-2012, 11:57 AM   #1
Olivia16
Junior Member
 
Location: United States

Join Date: Feb 2012
Posts: 7
Post MiSeq Run Fail - Microsoft Issue?

Has anyone experienced their MiSeq run just failing with no warning?

We started a run on our MiSeq, everything looked good as far as cluster density, Q scores, etc. then in the middle of the night at cycle 169 the machine essentially turned off. When I looked at the machine the next morning the computer was not responding. When the machine was turned back on and I went to the Illumina interface it just went to the main home screen page as if you were about to start a run. When I hit sequence the machine told me I forgot to do a post-run wash and acted like it had completed the run (no error messages were generated). I called Illumina and their first thought was that this was due to a glitch on the Microsoft side. After some of their reps looked through all our files they couldn't find any reason why the run would fail on the Illumina side of things. However when looking at the Microsoft side there were several error messages generated.

We are having a tech come out and look at our machine this week, but I just wanted to see if anyone else had experienced their MiSeq run failing due to glitches in the Microsoft software? Illumina didn't portray that this has been an issue in other MiSeq machines but I just want to get an idea if anyone else has had a similar issue.

Thanks for the help!
Olivia16 is offline   Reply With Quote
Old 05-19-2012, 12:07 AM   #2
beardy_man
Junior Member
 
Location: UK

Join Date: Mar 2011
Posts: 4
Default

Yes we have just had the same issue. We got about 100 cycles into a run then the software spontaneously crashed. All we got was a MS error box and no data at all. We have been told we need a hard drive replacement. We also contacted a number of colleagues any found 4 similarly broken MiSeqs out there.
beardy_man is offline   Reply With Quote
Old 06-04-2012, 12:56 AM   #3
Markiyan
Senior Member
 
Location: Cambridge

Join Date: Sep 2010
Posts: 116
Exclamation Putative miSEQ embedded PC system hardware/software design faults.

Did you check your's computer board's hardware monitor voltages? Did you have automatic updates enabled?
On ours Instrument we had a few issues with the slow performance of the computer, it turned out to be in the following configuration:
A. Hardware issues.
The power supply voltages to the (embedded PC) motherboard are (according to aida64):
+12V rail = +9.7V
+ 5V rail = +5.7V
+3.3V rail = +3.4V
-5V rail = (somewhere between -1.2V - -2.5V),
-12V rail = (similar to -5V rail, way bellow -12V).
So we clearly have a M/B PSU problem here. (+12V rail overload or failure) It's really surprising that it still was working at all... The support seemed to be unaware of this issue (but they knew about the HDD failures, which were probably caused by +5V going up even more under the load to +12V rail)...
Also as for the HDD it uses a standard 2.5' SATA laptop HDD
which will be getting 5.7 or more volts instead of 4.7-5.5V.

Illumina's R&D and support are in the definite need of a donation of the few +12V Halogen/car bulbs for testing of their's embedded systems PSU's under the load (use them for +12, +5V, +3.3v rails + a few multimeters)...
(Some PSU stuff made around the pacific region is able to provide only 1/2 - 1/3 of the rated power/current).
Also use infrared thermometer/FLIR camera for checking the thermal modes of all components under the full load conditions. The semiconductors doesn't like temperatures above 60 degrees (esp. power regulators/MOFSETS).

B. Software issues.
1. Automatic updates were enabled.
2. Windows defender switched on.
3. Illumina's antivirus was on (resets all changes (including network configuration) on the drive C on the reboot).
All of the above was switched off. It worked somehow until they had tried replacing the HDD (only the second replacement attempt was partially successful), then we started getting interrupted runs after 40-70 cycles (saying: "communication error") - TCP/UDP timeout?...

PS: I would really like to know who/(which company) exactly was responsible for embedded PC specifications setting, assembly, hardware/software configuration and QC. Looks like it probably was outsourced and it was their first PC they had ever assembled/configured in their life (did they move their plants from China to Malasia recently?)... They seem to have absolutely no idea about proper instrument control PC hardware/software (OS) configuration.

PPS: If the other miSEQ owners could post their PC motherboard's readings (esp. +5V and +12V rails), it would be helpful in estimating the scale of the problem.
Markiyan is offline   Reply With Quote
Old 06-05-2012, 12:41 AM   #4
simonandrews
Simon Andrews
 
Location: Babraham Inst, Cambridge, UK

Join Date: May 2009
Posts: 871
Default

On a related note, has anyone else had problems with their MiSeqs flooding their pipeline server with Samba related error messages? We're pushing MiSeq data through the same pipeline as our HiSeq and GA, but the MiSeq nearly shut down our pipeline server by filling up /var/log/messages with Samba errors. One of these is a known issue where Windows requests a share name with the last letter removed, which I've fixed, but it's still making unauthenticated requests to the samba server. These aren't related to the mounted drive we use to export data, which is working fine and seems to be the MiSeq trying to connect to the pipeline server as a domain controller.
simonandrews is offline   Reply With Quote
Old 06-06-2012, 01:57 PM   #5
Markiyan
Senior Member
 
Location: Cambridge

Join Date: Sep 2010
Posts: 116
Default win7/samba Master browser heated election campaighn consequences...

Quote:
Originally Posted by simonandrews View Post
These aren't related to the mounted drive we use to export data, which is working fine and seems to be the MiSeq trying to connect to the pipeline server as a domain controller.
In this case it looks like there is some "heated discussion" on the network, who should be "master browser" - so set both miseq windows 7 and your samba not to be master browser.

http://www.google.com/search?q=disab...wser+windows+7
http://www.hightechdad.com/2007/05/0...d-8003-errors/
Markiyan is offline   Reply With Quote
Old 06-09-2012, 01:12 PM   #6
pmiguel
Senior Member
 
Location: Purdue University, West Lafayette, Indiana

Join Date: Aug 2008
Posts: 2,317
Default

Quote:
Originally Posted by Olivia16 View Post
Has anyone experienced their MiSeq run just failing with no warning?

We started a run on our MiSeq, everything looked good as far as cluster density, Q scores, etc. then in the middle of the night at cycle 169 the machine essentially turned off. When I looked at the machine the next morning the computer was not responding. When the machine was turned back on and I went to the Illumina interface it just went to the main home screen page as if you were about to start a run. When I hit sequence the machine told me I forgot to do a post-run wash and acted like it had completed the run (no error messages were generated). I called Illumina and their first thought was that this was due to a glitch on the Microsoft side. After some of their reps looked through all our files they couldn't find any reason why the run would fail on the Illumina side of things. However when looking at the Microsoft side there were several error messages generated.

We are having a tech come out and look at our machine this week, but I just wanted to see if anyone else had experienced their MiSeq run failing due to glitches in the Microsoft software? Illumina didn't portray that this has been an issue in other MiSeq machines but I just want to get an idea if anyone else has had a similar issue.

Thanks for the help!
We just had something very similar happen last night. Run started, looks great for 120 cycles, then no more data seems to be collected. Went in to look at the instrument and see a Windows dialog box warning me that I am low on memory and offering to shutdown a program to free up memory for me. Also there is a much more cryptic MiSeq error window visible with hex addresses for memory locations. Although the MiSeq Constrol Software (MCS) is more-or-less hung, Windows is responsive, if a bit flakey.

I bring up task manager and indeed MCS looks to be hogging nearly all the system memory >1.5 gb and Real Time Analysis (RTA) is using 384 kb. So I am thinking "memory leak"?

Anyway, I decline Window's offer to shut down some of the programs. But when I click the "ok" button on the MCS error message, MCS closes. Task manager shows all the memory returned to the system. So, my impression is that Windows is not the problem here, MCS is.

When I relaunch MCS, it "reinitializes" the instrument and makes no mention of the previous run. When I push the "sequence" button I am warned that the previous run did not do a wash and would I like to do one now. I don't see any offer the continue the run--probably would be hosed now anyway. Sniffing back tears at our lost run, I assent and swap out the reagent cassette and PR2 bottle for the water wash bottles.

So, anyway, the HiSeq Control Software is very "re-entrant". One gets the feeling you could shut down the computer for a week in the middle of run but when you brought it back up you would be able to continue the run. The MiSeq seems to be more of the "forget the past, move on" philosophy.

--
Phillip
pmiguel is offline   Reply With Quote
Old 06-11-2012, 07:30 AM   #7
pmiguel
Senior Member
 
Location: Purdue University, West Lafayette, Indiana

Join Date: Aug 2008
Posts: 2,317
Default

Now that I look more closely, this doesn't look to be a memory leak. The low memory issues appear in the Windows logs hours after the primary event. The primary event (the event that occurred just prior to the instrument no longer running) was a .NET error:

Description:
Application: MiSeqControlSoftware.exe
Framework Version: v4.0.30319
Description: The process was terminated due to an internal error in the .NET Runtime at IP 000007FEF8EA929C (000007FEF8D90000) with exit code 80131506.


Any ideas?

--
Phillip
pmiguel is offline   Reply With Quote
Old 06-11-2012, 07:46 AM   #8
colindaven
Senior Member
 
Location: Germany

Join Date: Oct 2008
Posts: 415
Default

My bet is that the whole system moves to linux inside 12 months !
colindaven is offline   Reply With Quote
Old 06-12-2012, 12:47 AM   #9
simonandrews
Simon Andrews
 
Location: Babraham Inst, Cambridge, UK

Join Date: May 2009
Posts: 871
Default

Quote:
Originally Posted by colindaven View Post
My bet is that the whole system moves to linux inside 12 months !
I'd take that bet I can't see them changing the front end any time soon.

There's nothing inherently unstable about windows - our GAs and HiSeq have been very solid - it seems that this is more of a QC issue in the MiSeq software.
simonandrews is offline   Reply With Quote
Old 03-09-2015, 07:53 AM   #10
cement_head
Senior Member
 
Location: Oxford, Ohio

Join Date: Mar 2012
Posts: 253
Default

Hello All,

Just had a MiSeq Controller Software failure. Apparently on of the required files is missing/corrupted. Called Illumina, their engineer stated it was liklely a HDD failure and they will be oout in a couple of days to replace the onboard HDD.

This is the first hardware failure we've had in about 18 months of operation.

Regards,
Andor
cement_head is offline   Reply With Quote
Old 03-09-2015, 08:13 AM   #11
Markiyan
Senior Member
 
Location: Cambridge

Join Date: Sep 2010
Posts: 116
Default Miseq V2 seem to be much better quality (after 3 yeras of use :-)

Anyway, it looks like Illumina had responded well to the problems with V1 machine and had fixed major issues with in their Miseq V2, which has a beefier PC, and Nextseq, which finally uses RAID1 for their instrument disk subsystem.

With respect of improving the HDD lifetime expectancy in the MiSeq system, I would recommend doing a few things:

0. Backup entire HDD using Acronis or Paragon or Clonezilla imaging soft.
1. Enable microsoft AHCI driver in windows;
2. Turn on AHCI in the CMOS setup (BIOS) - switch from the IDE to AHCI,
(if any BSOD during windows bootup, make sure step 1 is done properly (revert to IDE mode for diagnostics);
3. Disable buffer flushing for futher performance improvement;
4. Backup all data from D:\ partition and make sure to format it with NTFS 64KB clusters!, than restore data back;
5. Anvanced users can also converrt cluster size to 64KB on the C: drive, and do a defragmentation after it. (Using Acronis or Paragon liveUSB, make sure to have a backup, just in case). If any boot problems use BOOTIce from live PE usb to fix;

The above steps greatly reduce the stress on mechanical HDD's, and extend their life.

6. (Optional) Put in an enterprise grade 1TB+ MLC SSD in and restore the image to it :-)
Markiyan is offline   Reply With Quote
Reply

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off




All times are GMT -8. The time now is 07:44 PM.


Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2020, vBulletin Solutions, Inc.
Single Sign On provided by vBSSO