Server upgrade
log in

Advanced search

Message boards : News : Server upgrade

1 · 2 · Next
Author Message
Profile Kyong
Project administrator
Project developer
Project tester
Project scientist
Avatar
Send message
Joined: 9 Jun 12
Posts: 532
Credit: 51,029,904
RAC: 4,962
Message 2436 - Posted: 24 Jan 2014, 11:33:36 UTC

There is going to be another server upgrade on Monday 27.01.2014 starting at 07:30 UTC. RAID array and filesystem will have to be adjusted so it will take more time. Server will be offline approximately 3 days. I assume being online on Wednesday 29.01.2014 till 11:00 UTC. If there will be more delay I will inform about it on twitter.

Radim VanĨo (Kyong)

Profile tom richardson
Avatar
Send message
Joined: 8 Sep 13
Posts: 40
Credit: 3,233,640
RAC: 2,345
Message 2437 - Posted: 24 Jan 2014, 12:42:00 UTC - in response to Message 2436.

Many thanks for the info, just increased my buffer to accept more work before the upgrade to keep my computer busy while the upgrade takes place.

Best regards,

Tom.

tdkmatt
Send message
Joined: 5 Aug 13
Posts: 2
Credit: 432,840
RAC: 0
Message 2461 - Posted: 30 Jan 2014, 5:37:54 UTC

Hi Kyong

i have a very powerful home datacenter and quite a bit of free storage in my array i can spare, and any addition help server wise

Profile Kyong
Project administrator
Project developer
Project tester
Project scientist
Avatar
Send message
Joined: 9 Jun 12
Posts: 532
Credit: 51,029,904
RAC: 4,962
Message 2463 - Posted: 30 Jan 2014, 8:59:40 UTC

Hi, thanks, but we have enough space for now.

MarkJ
Avatar
Send message
Joined: 27 Jun 12
Posts: 104
Credit: 35,768,960
RAC: 4,719
Message 2464 - Posted: 30 Jan 2014, 10:28:54 UTC - in response to Message 2463.

Hi, thanks, but we have enough space for now.

Did you add an extra drive to the RAID array? Given the drives are fairly small capacity (only 600Gb) I would think you would need quite a few drives. How many can you add to the server and how much do they cost?

I ask because it might be an idea to add as many as possible at once to save having to rebuild the RAID array each time, which can take days at a time.
____________
BOINC blog

Profile Kyong
Project administrator
Project developer
Project tester
Project scientist
Avatar
Send message
Joined: 9 Jun 12
Posts: 532
Credit: 51,029,904
RAC: 4,962
Message 2466 - Posted: 30 Jan 2014, 15:42:25 UTC

Yes, I added other two hard drives, so now we have 3.3 TB available and 2 x 8 GB RAM was also added. We have currently 1.3 TB used. One hard drive costs about $500 and there is 7 of 8 hard drives in the server. So there just one slot left. I was adding them by two but the RAID controlled didn't allow me to add more than one hard drive at once to the array.
There are also no all data we have, we would need about 5 TB just for data, but we also have to process all results so there is no need to have more capacity for one server because of load. I suppose that later the load will be higher and we buy a new server for database and our backup computer have pci-e RAID controller with 3 SAS driver so we add new hard drivers later there and connect it via sshfs, so the other processing will use resources of another server with additional capacity.

Prusas
Send message
Joined: 23 Jan 14
Posts: 1
Credit: 65,280
RAC: 0
Message 2469 - Posted: 30 Jan 2014, 17:57:16 UTC

Why not schedule a video?

Profile Overtonesinger
Avatar
Send message
Joined: 9 Sep 13
Posts: 17
Credit: 11,173,320
RAC: 11,235
Message 2471 - Posted: 30 Jan 2014, 21:26:59 UTC - in response to Message 2466.

WoW. 7! Thats a lot of disks. :)

Please, Kyong, do You know of anyone who has programmed a boinc screensaver?
I am really thinking about (after the OpenCL version) the possibility to show the data - what is exactly being computed. I mean - "morphing asteroid" probably, because it is searching for the shape, right?
And light from the right side ..., the current rotation period, etc.
____________
Melwen - child of the Fangorn Forest

tdkmatt
Send message
Joined: 5 Aug 13
Posts: 2
Credit: 432,840
RAC: 0
Message 2473 - Posted: 31 Jan 2014, 3:43:38 UTC

What OS are you using?

Dagorath
Send message
Joined: 16 Aug 12
Posts: 293
Credit: 1,116,280
RAC: 0
Message 2475 - Posted: 31 Jan 2014, 5:25:16 UTC - in response to Message 2473.

What OS are you using?


For reliability, robustness, user friendliness and TCO, probably Linux and I would bet Debian Linux.

@Kyong,

I almost fell out of my chair when you said $500 for a 600 GB hard drive. Then I checked the server status page where it says they are IBM 600 GB 10,000 rpm 6 Gbps SAS. Do you have the 2.5" hybrid drives described in the link?
____________
BOINC FAQ Service
Official BOINC wiki
Installing BOINC on Linux

Profile HA-SOFT, s.r.o.
Project developer
Project tester
Send message
Joined: 21 Dec 12
Posts: 173
Credit: 45,906,240
RAC: 56,740
Message 2477 - Posted: 31 Jan 2014, 11:50:37 UTC - in response to Message 2475.
Last modified: 31 Jan 2014, 11:57:40 UTC

Do you have the 2.5" hybrid drives described in the link?


I think it's not hybrid drive. It's 10k high speed server disk. These disks are more expensive.

Profile Kyong
Project administrator
Project developer
Project tester
Project scientist
Avatar
Send message
Joined: 9 Jun 12
Posts: 532
Credit: 51,029,904
RAC: 4,962
Message 2494 - Posted: 3 Feb 2014, 17:17:51 UTC

Yes, they are not hybrid disks, just high speed.

Profile mikey
Avatar
Send message
Joined: 1 Jan 14
Posts: 253
Credit: 6,581,040
RAC: 0
Message 2496 - Posted: 3 Feb 2014, 17:25:34 UTC - in response to Message 2494.

Yes, they are not hybrid disks, just high speed.


Don't hybrid drives work thru a cache system? If so they may not work very well if your system crashes mid write, so depending on the cache size they could make things less reliable then the expensive ones.

Dagorath
Send message
Joined: 16 Aug 12
Posts: 293
Credit: 1,116,280
RAC: 0
Message 2498 - Posted: 4 Feb 2014, 3:28:13 UTC - in response to Message 2496.

Not 100% sure but I think all disks including SSD use a volatile cache/buffer. In addition the OS itself has (or can have) a volatile cache/buffer.

The better (say more expensive) hardware RAID controllers often have a volatile cache (so it's fast) that is on an on-card battery backup that can maintain the cache content for a few hours, maybe more, hopefully long enough to allow the sysadmin to fix the problem, at which point the cache gets flushed to disk. That adds a layer of reliability. Not sure if A@H has such a RAID card, they are expensive.

I think maybe there are also ethernet NICs that have a battery maintained cache too? To prevent losing data blocks that have been acknowledged but not saved to non-volatile storage?

A reliable server also has a UPS and software that does a graceful shutdown if the power goes off and stays off for more than x minutes. I have a UPS on one of my rigs and I am convinced that of all the factors that help make that rig stable and robust, the UPS is the biggest factor. It's a Linux machine which is pretty stable compared to Windows but some power failures mess things up anyway. The worst are the ones where the power stutters (goes on and off or else down and back up very quickly) a few times and the rig doesn't die immediately. Those are the worst! A UPS prevents all of that.
____________
BOINC FAQ Service
Official BOINC wiki
Installing BOINC on Linux

Profile mikey
Avatar
Send message
Joined: 1 Jan 14
Posts: 253
Credit: 6,581,040
RAC: 0
Message 2499 - Posted: 4 Feb 2014, 13:58:25 UTC - in response to Message 2498.

Not 100% sure but I think all disks including SSD use a volatile cache/buffer. In addition the OS itself has (or can have) a volatile cache/buffer.

The better (say more expensive) hardware RAID controllers often have a volatile cache (so it's fast) that is on an on-card battery backup that can maintain the cache content for a few hours, maybe more, hopefully long enough to allow the sysadmin to fix the problem, at which point the cache gets flushed to disk. That adds a layer of reliability. Not sure if A@H has such a RAID card, they are expensive.

I think maybe there are also ethernet NICs that have a battery maintained cache too? To prevent losing data blocks that have been acknowledged but not saved to non-volatile storage?

A reliable server also has a UPS and software that does a graceful shutdown if the power goes off and stays off for more than x minutes. I have a UPS on one of my rigs and I am convinced that of all the factors that help make that rig stable and robust, the UPS is the biggest factor. It's a Linux machine which is pretty stable compared to Windows but some power failures mess things up anyway. The worst are the ones where the power stutters (goes on and off or else down and back up very quickly) a few times and the rig doesn't die immediately. Those are the worst! A UPS prevents all of that.


Thanks for the education. As far as UPS's I have several over the years but the batteries in them get expensive when you have more then just a couple and I have gone to just suing the surge protector side of them, but also now have a whole house surge protector as well. If you live in a house I recommend you check them out, mine was professionally installed and was only about 150 bucks US. I had the guy do it when he added some dedicated circuits for my pc's.

Qivis
Send message
Joined: 12 May 13
Posts: 1
Credit: 111,720
RAC: 0
Message 2523 - Posted: 14 Feb 2014, 19:30:43 UTC

Nothing has been available for my computer to work on for this project since late January - just before the server update. According to the server status though, the server is running just fine. And except for adding a few days more buffering to the BOINC settings - and as it turns out a few hours before the server went offline for the upgrade - I have not changed anything in the settings.

Profile mikey
Avatar
Send message
Joined: 1 Jan 14
Posts: 253
Credit: 6,581,040
RAC: 0
Message 2525 - Posted: 15 Feb 2014, 12:55:46 UTC - in response to Message 2523.

Nothing has been available for my computer to work on for this project since late January - just before the server update. According to the server status though, the server is running just fine. And except for adding a few days more buffering to the BOINC settings - and as it turns out a few hours before the server went offline for the upgrade - I have not changed anything in the settings.


I think you have something misconfigured on your screen then because you DO have units finishing, like this one:
28789133 11560066 12 Jan 2014, 19:14:59 UTC 14 Jan 2014, 1:51:17 UTC Completed and validated 10,750.39 100.32 480.00 Period Search Application v101.00 (cuda55)

that was received by you on 12 Jan 2014, 19:14:59 UTC and returned for credit on 14 Jan 2014, 1:51:17 UTC. You also have 46 units in progress right this minute. You also got this unit:
27956701 11964914 5 Jan 2014, 21:26:51 UTC 6 Jan 2014, 21:44:19 UTC Completed and validated 10,733.23 103.91 480.00 Period Search Application v101.00 (cuda55)

On 5 Jan and returned it for credit on 6 Jan. In short your computer seems to be doing just fine getting and returning valid units with no problems. You are talking about the AMD quad core with the Nvidia GTX 650 in it right? Because that is the only pc listed on your account, if you have another maybe it somehow got detached and you need to reset it.

Andrew Davis
Send message
Joined: 8 Aug 13
Posts: 46
Credit: 1,503,000
RAC: 0
Message 2527 - Posted: 15 Feb 2014, 15:29:32 UTC

Is there any problem transferring work-units from 7.2.33 to 7.2.39?

swiftmallard
Avatar
Send message
Joined: 12 Jan 13
Posts: 38
Credit: 4,421,520
RAC: 0
Message 2528 - Posted: 15 Feb 2014, 16:52:05 UTC

No, just shut down Boinc and install the new version over the old. You'll lose nothing.

Manning M Widener
Send message
Joined: 30 Jun 13
Posts: 7
Credit: 4,790,640
RAC: 0
Message 2575 - Posted: 22 Feb 2014, 16:53:46 UTC

Does a problem exist with the data server or has all available asteroid data been distributed for analysis? Have 5 tasks "Ready to report" on one laptop and no tasks are available for another computer to work on. Thanks....

1 · 2 · Next
Post to thread

Message boards : News : Server upgrade


Main page · Your account · Message boards


Copyright © 2017 Asteroids@home