4 ms·

Having worked with Nokia Networks as a customer on Wireless core networks, I don't believe some of these conclusions are safe. Although in concept I agree, the

Having worked with Nokia Networks as a customer on Wireless core networks, I don't believe some of these conclusions are safe. Although in concept I agree, they are and will continue to be a major player for the foreseeable future, I suspect some of your other conclusions may be conflating separate entities. I don't know about area's becoming ghost towns, but I do know some areas have been impacted, and may be coming back with a startup culture that is up and coming. Overall I don't have good insight though. We had professional services come over from Finland in many instances, but as I recall all the product lines we were using were developed in other european countries and India. After the Alcatel-Lucent merger, they're also picking and choosing products to keep and killing off teams from what I've heard (or in some cases trying to merge products). I don't think it's safe to assume what was sold to Microsoft carries over to other business units. The remaining networks business units were not part of what got sold off, and was already a conglomerate of mergers and acquisitions. I'm sure all business units have some compelling products however. And I won't comment on the Finns or their country, I have absolutely no insight. My experience with Nokia Networks equipment though (only indirect experience with RAN equipment), is in general it's not something I would want to choose to buy. Stood up against their competitors though with similarly rough solutions, it might still be the winner, but I know there will be lots of specific pain involved with Nokia's wireless products. The docs are rough, and things like upgrades are not really meant to be run by customers (Nokia commonly runs the networks for their customers, we had a different and more hands on culture). The configuration of the system is effectively managed by giant excel documents and obscure tools, leading to lots of lost/missing configuration updates, and a lack of source of authority in configuration data (upgrades use excel, running system uses local data). I'm slightly out of date and they've been promising to get rid of this forever, but that was the state of affairs a couple years ago and the last rumour I heard is it wasn't gone yet. With some of their competitors, we solved really obscure bugs, problems that affected 0.000001% of customers per day, with Nokia it was just so much harder and longer to solve any issues, I wouldn't even dream of being able to solve the rare problems. What Nokia was good at, is when the product was running, it was fairly stable, so while operating it wasn't fun, the parts I worked with were stable, and seemed relatively well designed and high performance. With the mergers in the telecom industry, if you want one or just a few vendors, Nokia is a competent choice that you can build your entire network off of. The products I really enjoyed working with in the space though, none of them came from Nokia. * My thoughts are my own, and do not reflect those of any current/former employer.

iagovar 6y ago
Which ones did you enjoyed? I work for an euro telco and we have vendors all over the place, nokia, ericsson, huawei... I work with some (many) software abstraction layers but the few times I had to interact with equipment directly I found Alcatel to be the most difficult. Anyway the whole space looks obscure and full of propietary solutions.
- kevin_nisbet 6y ago
  In it's heyday Starent equipment was excellent, but I don't think it's as good since Cisco bought them. So for MME's, S/P-GW's, etc I would use again. This was the standard that I tried to hold the other vendors to on support, upgradeability, configuration, etc. The OAM / support systems were pretty weak, but our team just didn't bother to use them, and the built in tools were better. For DNS, infoblox appliances were good. Infoblox didn't have much experience with the GSMA DNS setup or carrier roaming, so they weren't well positioned to help us, but most the problems we had were with the GSMA specs and our partners being wonky, not Infoblox itself, so we managed to get through those issues on our own. I never used in production, but we had Acme packet SBC's in our lab for awhile, which looked like they'd be great. But that's under Oracle's umbrella now, so probably something to avoid. Our teams also really liked Tekelec for SS7 routing but I never used it myself. I looked at Tekelec's DRA and it looked like it was pretty good and well designed, but again it's Oracle now. I don't have much direct experience with alcatel's stuff in production. I worked with them on a reseller agreement, so the resold equipment doesn't really reflect them, and doing some evals. Alcatel's CSCFs looked competent, but I think our PCRF team was really struggling from what I saw. And yea, telecom is in that state, were it's jointly developed standards, but the implementations are very vendor driven. As an industry, the software all needs to be high performance, high capacity, and it's following millions of pages of standards, and then the add on's that the carriers want to do. I don't think there is much incentive for telecoms to go OSS on the telecom handling, although multiple vendors are starting to build off of OSS cloud native stuff. So it'll be interesting to see providers using kubernetes for example for a control plane and upgrades / capacity expansion. Or whether it's possible to integrate with an open tracing. In my view, OSS for along time was way behind what the telco's were doing from a troubleshooting / high availability / services perspective. But the establishment of the CNCF and all the projects under that umbrella seem to be jumping over what telco's had as proprietary solutions. How those get merged in will be interesting to see.

iagovar 6y ago

Which ones did you enjoyed? I work for an euro telco and we have vendors all over the place, nokia, ericsson, huawei... I work with some (many) software abstraction layers but the few times I had to interact with equipment directly I found Alcatel to be the most difficult. Anyway the whole space looks obscure and full of propietary solutions.

kevin_nisbet 6y ago
In it's heyday Starent equipment was excellent, but I don't think it's as good since Cisco bought them. So for MME's, S/P-GW's, etc I would use again. This was the standard that I tried to hold the other vendors to on support, upgradeability, configuration, etc. The OAM / support systems were pretty weak, but our team just didn't bother to use them, and the built in tools were better. For DNS, infoblox appliances were good. Infoblox didn't have much experience with the GSMA DNS setup or carrier roaming, so they weren't well positioned to help us, but most the problems we had were with the GSMA specs and our partners being wonky, not Infoblox itself, so we managed to get through those issues on our own. I never used in production, but we had Acme packet SBC's in our lab for awhile, which looked like they'd be great. But that's under Oracle's umbrella now, so probably something to avoid. Our teams also really liked Tekelec for SS7 routing but I never used it myself. I looked at Tekelec's DRA and it looked like it was pretty good and well designed, but again it's Oracle now. I don't have much direct experience with alcatel's stuff in production. I worked with them on a reseller agreement, so the resold equipment doesn't really reflect them, and doing some evals. Alcatel's CSCFs looked competent, but I think our PCRF team was really struggling from what I saw. And yea, telecom is in that state, were it's jointly developed standards, but the implementations are very vendor driven. As an industry, the software all needs to be high performance, high capacity, and it's following millions of pages of standards, and then the add on's that the carriers want to do. I don't think there is much incentive for telecoms to go OSS on the telecom handling, although multiple vendors are starting to build off of OSS cloud native stuff. So it'll be interesting to see providers using kubernetes for example for a control plane and upgrades / capacity expansion. Or whether it's possible to integrate with an open tracing. In my view, OSS for along time was way behind what the telco's were doing from a troubleshooting / high availability / services perspective. But the establishment of the CNCF and all the projects under that umbrella seem to be jumping over what telco's had as proprietary solutions. How those get merged in will be interesting to see.