Service and Support
User experiences with IBM customer service indicate positive feedback. Technical personnel often communicate with IBM, receiving effective assistance for user issues. Support is described as knowledgeable, responsive, and helpful, offering quick problem resolution and accommodating gestures like demo software access. While some users rarely engage with IBM, reports suggest swift help during emergencies and a strong customer service reputation, particularly in complex projects, like those with a leading German bank.
Deployment
Several users describe IBM Spectrum Scale's initial setup as complex, especially for large clusters requiring preparation and documentation. Some find it straightforward with familiarity, while Red Hat implementation presents challenges. Deployment duration varies from a few days to a month, often depending on external factors like customer meetings. Vendor support simplifies the process for some. Installation phases include quick setup followed by a longer testing period, influenced by factors like network advisor deployment and avoidance of erasure coding.
Scalability
IBM Spectrum Scale demonstrates impressive scalability, though some users believe it has reached its growth potential. It allows expansion by adding more storage and servers, supporting a wide range of clients from small businesses to large banks. Users highlight its robustness, unmatched by other systems like NFS or HACMP. Despite requiring substantial resources, it is praised for resilience during network interruptions, with scalability consistently rated highly across varied deployments.
Stability
IBM Spectrum Scale is generally stable, with users noting reliability and minimal issues. Upgradability presents challenges, as updates require time for each node. Despite occasional server problems, the cluster design prevents significant disruptions. Many users describe it as incredibly stable, with no turmoil experienced even with intense workloads. Ratings are high, with some users giving it a perfect score. Erasure coding could pose minor slowdowns, but stability remains strong.