I know I know this is a Digium site but I just need to know. What cards do you guys use? I currently am testing with a TDM22B card and it is okay. I am debating getting a Sangoma card to also test with. What do you guys think about the Sangoma’s? Looking at prices, they look around about the same cost down here… So that isnt really a factor. Do the Sangoma cards work better in regards to Echo Cancellation and IRQ conflicts?
What hardware do you use? Motherboard recommendations, brands,models? Something that also intrigues me is that Sangoma have brought out the PCI Express cards… PCI slots are coming harder to come by. I know some PCI Express slots are backward compatiable with PCI cards. Anybody using a PCI VOIP card in a PCI Express slot?
I’ve purchased Sangoma cards for my installs for the IRQ/motherboard peace of mind. They claim their cards will work on any motherboard. And that multiple cards in the same box won’t cause IRQ conflicts. In my limited experience this has proven to be true.
I have been running a TDM400 card with no problems in an old dell desktop. I tried to migrate it to my samba/email/web server (running same 2.6.9-42 kernel - redhat based) on an intel entry-level motherboard. no joy (and i wasted HOURS on this…) in one pci slot, it’s recognized as a network controller in the other, it is recognized correctly, but doesn’t work. at all. one fxo and one fxs port. outside calls: ring no answer. inside calls: no dialtone even. no indication from asterisk log that the board is even seeing any requests from either the fxo or fxs daughtercards. i gave up
you’ll get all sorts of answers…some guys like HP, others like Dell, others like whatever is cheap.
as for the Sangoma/Digium debate, I’m partial to Sangoma, but Digium has worked pretty well for us overall. Sangoma cards seem to be slightly more forgiving in regards to hardware compatibility, but there is nothing wrong with Digium if you’re careful.
More important is proper linux config - making sure you’re not running non-essential services, making sure X-windows and the framebuffer are disabled, etc…
i know this is a Digium forum, but i prefer Sangoma hardware so far. my experiences with the AFT-A200 range are better than with the TDM400, and i’ve had better performance with the AFT-A101u than with the Digium equivalents.
i don’t keep records of what hardware i’ve installed them in though … Mac G3s, Asus, Gigabyte and Intel mainboards, Dell and Compaq servers.
we have 8 poweredge 2850’s - other than some stability issues specific to the digium TE410P cards, they’re pretty nice servers. i am liking the RAID controllers less and less, but they’re fast and fairly inexpensive.
i’ve heard good things about the HP proliant series of servers - haven’t ever played with one, though
nothing much, other than the fact that three servers now have had a drive fail in RAID5 and the entire array is lost - it’d be one thing if it kept chugging along in read-only mode, but the systems have just hung…and getting the arrays rebuilt has only worked one out of the three times - otherwise we’ve had to start over fresh.
on our asterisk boxes, we’re running RAID1 with a hotspare, and that works great - we can pull the hotspare and one of the primaries and the system stays up - we’d pretty much have to lose all three drives for the system to fail. you lose quite a bit of space this way (3:1 storage ratio) but it’s VERY fault tolerant.
maybe it’s just me, but i thought enterprise RAID controllers were supposed to be a bit hardier…
I also have Dell System - all the servers here are dell - and have had no problems with their raid controllers - Have had a couple drives in our servers DIE, and just popped in a new drive and it rebuilt the array - all of our setups are with atleast 4drives in a raid 5 config - I have not expirence in a 3drive raid5 config with these raid controllers
all our arrays were 5 drives with a hotspare, except for our NAS box which was a 4 drive array, no hotspare.
two of the 2850’s failed when 1 drive failed, and the NAS server failed this morning actually. i can’t explain it, but i just have had bad luck when a drive fails.
to be fair, the NAS box didn’t lose the array…yet…but dell is stating that the server should be able to boot, but it can’t - the array is showing as FAILED in the controller BIOS, and they don’t know why.
glad to hear others are not having the issues we are…i’d love to hear what people think of the HP series though - heard a LOT of good things about those.
Interesting experance on Dell Raid5 controllers. If the whole array fails or go to write only when only one drive goes down it is not Raid5.
Raid5 should be able to handle the loss of a single drive and still be able to read and write, along with being able to rebuild the failed drive once a good one is in its place.
I have worked with Dell hardware Raid5 controllers (though four years ago) and we used 5 drives with no spare (you should not need a spare in Raid5) and on four seperate occassions we replaced a failed drive with no loss of data or down time. Though when the new drive was inserted and being built, disk response is slower for reads and writes.
Something is not right with your “RAID5” since it is not acting and behaving as a RAID5 should.
reason we went with a hotspare was so that in the event of a drive failure, the array would automatically “insert” the spare drive, so that we would actually have to lose two drives for there to be any real danger of data loss.
i agree with you all that raid 5 SHOULD have worked, but for some reason (perhaps something unique to the situations in which the array failed that either escaped my notice or that i’m just forgetting about now) we’ve had bad luck.
I’m looking forward implementing an Asterisk PBX and was looking for dell server compatibility with Digium Cards (SC1425, I saw SC1420 has issues).
Do you guys with RAID /disk issues have Maxtor SCSI drives, either 10k or 15k RPM models? A critical firmware update that addresses exactly this on these drives has been released recently…
When you have a new server, with a RAID setup always test a data recovery/ raid failure!
Install an os do some things and pull out one of the hotswap drives…
Re-insert it and make sure it rebuilds! If it doesnt then you have problems.
Do you check your backups? Have you ever done a FULL data recovory? This should be number one on things to do before a machine goes into a production environment.
in the other, it is recognized correctly, but doesn’t work. at all. one fxo and one fxs port. outside calls: ring no answer. inside calls: no dialtone even. no indication from asterisk log that the board is even seeing any requests from either the fxo or fxs daughtercards. i gave up